Most Dangerous Countries for Women: Mapping Femicide

The Most Dangerous Countries for Women: Mapping Femicide

The story that finally compelled me to write this article was the discovery of the bodies of nine women in Ekurhuleni, east of Johannesburg, over the course of approximately two months. Some had reportedly been found naked or partially clothed; some appeared to have suffered sexual violence. The similarities between the deaths led to understandable fear that a serial killer might be targeting women in the area, although investigators had not established that the murders were the work of one offender. By 19 September 2026, South Africans were gathering for memorial runs and protests, mourning the women while demanding answers from a country that has lived with exceptionally high levels of gender-based violence for far too long. (Reuters)

I found the story particularly difficult to put down because I have spent years researching femicide, intimate-partner violence and the conditions that allow women’s deaths to be treated as disconnected private tragedies. The possibility of a serial offender was horrifying, but so was the alternative. If the women had been killed by different men in unrelated circumstances, the appearance of nine bodies across one metropolitan region would not become less disturbing. It would reveal a form of danger so dispersed through ordinary life that no single monster could contain or explain it.

That thought led me back to a question that appears deceptively simple: where are the most dangerous places in the world to be a woman? It is the kind of question the internet usually answers with a numbered list, a composite index and a photograph of a woman in a non-Western country. Yet years of researching femicide have taught me that a single ranking is not merely inadequate; it can actively conceal the way danger works. The place in which women are most likely to be murdered by partners is not necessarily the place in which women face the most restrictive laws, the highest prevalence of sexual violence, the greatest risk of child marriage or the most extreme wartime violence. A country can recognise women’s formal equality and still fail to protect them from the men closest to them. Another can impose visible restrictions on women’s movement and public life while recording a lower female homicide rate than a formally liberal society elsewhere.

There is also a political reason to ask the question more carefully. Women’s rights are routinely used as evidence in wider arguments about religion, civilisation and the legitimacy of foreign states. Restrictions on dress, education, employment and freedom of movement in Afghanistan, Iran and elsewhere are serious abuses and should never be minimised. However, the international imagination of the “oppressed woman” is strikingly selective. It can identify a compulsory veil immediately while overlooking endemic partner violence in the Pacific, the disappearance and sexualised murder of women in Latin America, the killing of Indigenous women in wealthy states, dowry deaths in South Asia, or a woman terrorised inside a European suburban home by a man who has never accepted her right to leave him.

The question, therefore, is not whether women are genuinely oppressed in the societies most often singled out by Western political discourse. In many cases they unquestionably are. The more revealing question is this: what happens when we stop allowing geopolitics to define what danger to women is supposed to look like, and examine the evidence in all its uncomfortable complexity?

Mapping Femicide: Key facts

• Approximately 83,000 women and girls were intentionally killed worldwide in 2024, and an estimated 50,000 were killed by intimate partners or other family members—about 137 every day. (UN Women and UNODC)
• Africa had the highest estimated 2024 rate of intimate-partner and family-member femicide, at approximately 3 victims per 100,000 women and girls, followed by the Americas at 1.5, Oceania at 1.4, Asia at 0.7 and Europe at 0.5. (UN Women)
• The latest global prevalence estimates indicate that approximately 840 million women—31.6 per cent—have experienced physical or sexual intimate-partner violence, non-partner sexual violence, or both. (World Health Organization)
• The highest regional prevalence of intimate-partner violence is estimated in Oceania, excluding Australia and New Zealand; sub-Saharan Africa; and Southern Asia. (World Health Organization)
• At least 3,828 femicides, feminicides or gender-related killings of women were officially recorded across 26 Latin American and Caribbean countries and territories in 2024. (ECLAC)
• There is no completely comparable global country ranking for femicide. Legal definitions, police practices, forensic capacity, conflict, under-reporting and the quality of national statistics differ too greatly.

What does “the most dangerous country for women” actually mean?

Before identifying the most dangerous countries for women, we have to decide what kind of danger we are measuring. The probability that a woman will be murdered is one measure. The probability that she will be killed by a husband, boyfriend, former partner or relative is another. Rape and other forms of sexual violence, trafficking, forced marriage, female genital mutilation, reproductive coercion, economic exclusion, compulsory dress, restrictions on education and employment, conflict-related sexual violence and the practical impossibility of obtaining police protection all describe different ways in which a society can make women unsafe.

These risks overlap, but they do not produce one consistent global ranking. Afghanistan and South Africa can both be profoundly dangerous environments for women without being dangerous in the same way. Papua New Guinea cannot be assessed by the same measure as Honduras. The Democratic Republic of the Congo during armed conflict cannot be meaningfully placed beside France using a single number. India’s immense population, internal diversity and mixture of domestic abuse, dowry-related violence, sexual violence, son preference and abuse-related suicide cannot be compressed into one national position without losing most of what the data is trying to tell us.

This distinction is more than a methodological footnote. It changes the political meaning of the question. A ranking based on formal legal rights may identify countries where women cannot study, work, travel or participate freely in public life. A ranking based on intentional female homicide will produce a different map. A survey of intimate-partner violence will produce another, while an assessment of conflict-related sexual violence will necessarily concentrate on active war zones and displaced populations. None of those maps is false, but none is complete.

The search for one “worst country” also risks turning women’s suffering into an international contest in which one form of oppression is used to dismiss another. Acknowledging the systematic erasure of Afghan women from public life does not require us to ignore the woman murdered by a former partner in Britain. Recognising lower regional homicide rates in Europe does not require pretending that Europe has solved male violence. Nor should evidence of extreme violence in particular African countries be converted into an argument about a continent of 54 states as though it were one culture, government or social system.

The intellectually honest answer is that there are different maps of danger. The task is to place them over one another without allowing one to erase the rest.

Femicide: the fatal end of a much larger continuum

Femicide provides the thread through this wider landscape because it describes the point at which gendered danger becomes fatal. The term is sometimes used broadly to mean the killing of women because they are women, while researchers, campaigners and legal systems apply a range of more specific definitions. The UN statistical framework for measuring gender-related killings of women and girls defines the phenomenon through evidence of a gender-related motivation or dimension, including unequal power relations, gender stereotypes, discrimination, harmful norms and circumstances showing that the victim’s sex or gender was relevant to her death.

This means that female homicide and femicide are related but not interchangeable categories. A woman can be intentionally killed in circumstances where gender appears irrelevant. Conversely, a woman can be killed by a stranger, trafficker, family member, client, organised criminal group or rejected suitor in circumstances that are unmistakably gendered. Intimate-partner femicide is the most extensively measured form, but it is not the only one. Family-related femicide, sexual femicide, so-called honour killings, dowry-related deaths, killings connected to trafficking or sexual exploitation, and the murder of women accused of witchcraft can all sit within a wider femicide framework.

The latest joint estimates from UN Women and the United Nations Office on Drugs and Crime provide the clearest global baseline currently available. Of approximately 83,000 women and girls intentionally killed during 2024, an estimated 50,000 were killed by intimate partners or other family members. That is around 137 women and girls every day, or one every ten minutes. Approximately 60 per cent of female homicides therefore occurred within the supposedly private sphere, compared with about 11 per cent of male homicides. (UN Women and UNODC)

The difference matters. Men make up the majority of homicide victims worldwide, but men and women are not generally killed in the same circumstances or by the same people. Men’s greater overall exposure to homicide does not disprove femicide; it demonstrates why sex-disaggregated data and victim–perpetrator relationships are essential. A man is more likely to be killed in public violence by someone outside his family. A woman’s risk is far more likely to be concentrated in a home, relationship or family network that conventional ideas of “public safety” barely touch.

Femicide is therefore not the whole story of violence against women. It is the terminal point of a much longer story: the moment at which coercion, entitlement, sexual violence, harmful practices, state abandonment or organised brutality ends a woman’s life and eliminates every remaining possibility of escape.

Why there is no honest global femicide league table

Searches for “femicide rates by country” tend to produce neat tables with a number beside every state. The confidence of those rankings usually exceeds the confidence of the underlying evidence. Some countries record all intentional homicides of women; some identify intimate-partner killings; some combine partners and family members; and some count only cases prosecuted under a specific offence called femicide or feminicide. These datasets overlap, but they are not equivalent.

A woman sexually assaulted and murdered by a stranger may meet the international statistical definition of femicide but will never appear in an intimate-partner homicide dataset. A woman killed by her boyfriend can appear in partner-homicide data even if her country has no offence named femicide. A woman whose body is never found, never identified, or never properly investigated may not be classified at all. Conflict, weak forensic capacity, fragmented police systems and institutional indifference introduce further gaps.

The first dedicated international Statistical Framework for Measuring the Gender-related Killing of Women and Girls was endorsed only in 2022. Its purpose is to help states identify gender-related killings consistently regardless of the terminology used in domestic criminal law. The framework asks countries to collect contextual information about victims, perpetrators, prior violence, sexual violence, exploitation, harmful practices and institutional response. It exists because the simple fact that a homicide victim was female does not tell us whether her death was femicide, while a generic murder record can erase the evidence that it was.

There is also a perverse statistical effect: a country that creates a femicide offence, trains investigators and begins naming gender-related killings may appear to deteriorate, while a country that fails to investigate those same circumstances can look comparatively safe. Better recognition can increase the number recorded. Bad recognition can make violence disappear.

For that reason, this article does not present a definitive top ten. It examines where robust evidence reveals particular concentrations of danger and identifies what is being measured in each case. Absence from the data is not evidence of safety. Sometimes it is evidence that nobody counted her.

The world map changes according to what we measure

The global scale of non-fatal violence demonstrates why homicide data alone cannot answer the question. According to the World Health Organization’s latest violence-against-women estimates, approximately 840 million women and girls—31.6 per cent—have experienced physical or sexual intimate-partner violence, non-partner sexual violence, or both during their lifetimes. Among ever-partnered women, the estimated prevalence of physical or sexual partner violence is highest in Oceania excluding Australia and New Zealand, at 37 per cent, followed by sub-Saharan Africa at 32 per cent and Southern Asia at 31 per cent.

Those figures measure prevalence rather than lethality. They do not tell us that every woman exposed to partner violence faces the same risk of death, nor do they capture psychological abuse and coercive control adequately. Self-reported survey results are also shaped by disclosure, stigma, research access and the wording of questions. The WHO explicitly warns that the true prevalence of violence is likely to be higher than reported.

Nevertheless, the regional pattern is important because it exposes how much is omitted when public discussion focuses almost exclusively on the most politically visible restrictions imposed on women. Some of the highest documented levels of partner violence occur in countries that are rarely central to Western debates about women’s global oppression. Meanwhile, the forms of danger that attract the greatest geopolitical attention are not always those most likely to result in a woman being assaulted or killed by somebody she knows.

The map must therefore be read in layers. The first layer is lethal violence. The second is the wider prevalence of physical and sexual abuse. The third includes laws and state policies governing women’s movement, education, employment, dress, marriage and reproduction. The fourth reflects war, displacement and institutional collapse. The fifth asks whether police, courts and health services are capable of protecting women and holding offenders accountable. The most dangerous countries for women will change depending on which layer we examine, and the lives of real women are usually shaped by several layers at once.

Southern Africa: lethal violence and the story that prompted this article

Africa had the highest estimated regional rate of intimate-partner and family-member femicide in 2024: approximately 3 victims per 100,000 women and girls. The estimate represented about 22,600 deaths, but UN Women warns that major data gaps remain. The continent contains enormous variation, and continental figures should never be used to imply that women experience the same conditions in Botswana, Morocco, Mauritius, Nigeria and South Africa. They do, however, establish that the burden of lethal private-sphere violence is exceptionally high across parts of the region.

South Africa deserves particular attention because it possesses something many countries do not: more than two decades of dedicated national femicide research. The South African Medical Research Council has examined mortuary records, post-mortem reports and police investigations in order to identify relationships between victims and perpetrators, rather than relying only on cases legally labelled as femicide. Its fourth national study found that 60.1 per cent of women murdered in 2020/21 were killed by intimate partners, that approximately one in six murdered women showed evidence of sexual violence, and that fewer than one in five intimate-partner femicide cases resulted in a conviction. (South African Medical Research Council)

The country’s exceptional homicide burden exists within a wider environment of physical and sexual violence, inequality, violent crime, firearm access and institutional failure. It also exists alongside an immensely active women’s movement that has forced gender-based violence into national political debate. South Africa is therefore neither a convenient stereotype of African danger nor an inexplicable outlier. It is a country where researchers and campaigners have worked unusually hard to reveal the gendered patterns inside its homicide statistics—and where the state has repeatedly acknowledged a crisis it has not yet brought under control.

This is the context in which the Ekurhuleni deaths appeared. Police had found nine women’s bodies east of Johannesburg by 17 September 2026 and were investigating whether the cases were linked. The victims were reportedly young women; several were found naked or partially clothed, and similarities between the deaths caused alarm. Authorities had not confirmed the existence of a serial killer, and that uncertainty must be preserved. (Reuters)

Yet the meaning of the deaths is not suspended until investigators identify one offender or several. If one man killed multiple women, the case exposes the danger created by a predatory offender operating across a community. If different men killed the women separately, the case exposes a social environment in which several perpetrators could dispose of women’s bodies within the same region and the same brief period. The serial-killer frame attracts attention because it concentrates horror in one recognisable figure. Femicide research asks us to confront the possibility that the danger is often more ordinary, more distributed and more socially embedded than that.

Latin America and the Caribbean: disappearance, impunity and the politics of naming women’s deaths

No global account of femicide is complete without Latin America and the Caribbean. Feminist movements across the region helped transform femicidio and feminicidio from specialised concepts into legal, political and public language. The distinction between the terms varies by country and theorist, but feminicidio has often carried an important additional accusation: that the state’s failure to prevent, investigate and punish the killing is part of the phenomenon, not merely an unfortunate response to it.

The history of Ciudad Juárez in Mexico became internationally emblematic. From the 1990s onward, the disappearances and murders of women and girls exposed the convergence of sexualised violence, industrial border economies, organised crime, misogyny and impunity. Similar dynamics recur in different forms elsewhere in the region: intimate-partner killing, disappearances, attacks connected to criminal organisations, sexual violence, trafficking and ineffective justice systems. The result is not one uniform Latin American femicide pattern, but a regional political language capable of naming links that ordinary homicide categories had obscured.

According to the UN Economic Commission for Latin America and the Caribbean, at least 3,828 women were officially recorded as victims of femicide, feminicide or gender-related killing across 26 countries and territories in 2024—at least 11 every day. At least 19,254 such killings were registered over the five years to 2024. Among 17 Latin American countries reporting information for 2024, Honduras recorded the highest official rate, at 4.3 per 100,000 women, followed by Guatemala at 1.9 and the Dominican Republic at 1.5.

These numbers should not be dropped into a global table beside a female homicide rate from another country and treated as though they describe the same category. ECLAC itself cautions against simplistic comparison because reporting coverage and national definitions vary. Yet the data offers something indispensable: a regional record of killings that governments have been compelled to recognise as gender-related. By 2025, 20 Latin American countries had incorporated femicide, feminicide or gender-related killing into their legal systems, either as a specific offence or an aggravating circumstance.

The statistical paradox is unavoidable. Latin American countries can appear disproportionately visible on a femicide map partly because feminist movements succeeded in making the deaths visible. The appropriate conclusion is not that the region’s crisis is an artefact of counting; thousands of women are dead. It is that places without comparable recognition may be concealing deaths that would have been named as femicide elsewhere.

The region also shows why protest, memorialisation and naming matter. Ni Una Menos—Not One Woman Less—did not emerge because data alone created political change. It emerged because women refused the social ritual in which a murdered woman becomes a brief crime story, her behaviour is examined more closely than the perpetrator’s, and the structural pattern disappears before the next body is found.

The Pacific: one of the world’s most overlooked centres of violence against women

The Pacific is central to this article precisely because it is so often missing from popular conversations about the most dangerous places for women. WHO estimates place Oceania, excluding Australia and New Zealand, at the top of the regional table for lifetime physical or sexual intimate-partner violence among ever-partnered women. Small populations and limited data collection make country comparisons difficult, but national surveys in several Pacific states reveal extraordinarily high exposure to abuse.

Papua New Guinea provides one of the clearest examples. Its national demographic and health survey found that 58.3 per cent of ever-partnered women aged 15–49 had experienced physical or sexual intimate-partner violence at least once, while 47.6 per cent reported such violence during the preceding 12 months. (UN Women Data Hub) Those figures describe violence by a partner; they do not include every form of sexual violence, coercion, sorcery-accusation-related violence or abuse outside intimate relationships.

Studies from countries including Kiribati, Solomon Islands and Vanuatu have also reported high levels of partner violence, although survey years, age ranges and methods differ. Geography compounds the problem. Women living across remote islands or isolated rural areas can face severe barriers to medical treatment, policing, shelters and courts. A law against domestic violence provides little protection if the nearest service is inaccessible, the perpetrator controls transport and money, or customary and formal justice systems fail to respond.

Yet the women of the Pacific rarely function as the default symbol of global female oppression. Their relative invisibility exposes a central problem in the international conversation: visibility is shaped not only by severity, but by geopolitical usefulness, media infrastructure and the familiarity of the cultural symbol. A dress code is instantly photographable. Coercive control on a remote island is not. One becomes a recurring international headline; the other remains a prevalence figure buried in a survey.

South Asia: dowry, son preference and deaths hidden by category

South Asia contains another complex geography of danger. The region includes enormous differences of religion, class, caste, ethnicity, law and social structure, making broad cultural explanations particularly unreliable. However, certain patterns—including dowry-related abuse, so-called honour violence, child marriage, sexual violence, son preference and intimate-partner abuse—have generated distinct categories of gendered harm and death.

India demonstrates why a narrow homicide count cannot represent the full lethal continuum. Official data submitted by the Ministry of Home Affairs recorded 6,450 dowry deaths in 2022, alongside 4,963 cases registered as abetment to the suicide of women. The same table recorded 140,019 cases of cruelty by a husband or his relatives. (Government of India, Ministry of Home Affairs) These are police-recorded cases, not prevalence estimates, and each legal category has its own evidential limitations. They nevertheless show how gender-related deaths can be dispersed across murder, dowry-death and suicide classifications rather than appearing in one femicide total.

Dowry violence is particularly important because financial interest, family power and gender inequality can become inseparable. A bride may be harassed, assaulted or killed over demands for money or property; a death can be staged as an accident or suicide; and pressure can come from a husband, his relatives or a wider family system. The motive cannot be reduced to “money” as though it were an ordinary financial dispute. The economic demand is structured by a marital system in which a woman’s value, security and survival can be made conditional on transfers from her family.

Son preference and gender-biased sex selection extend the question of lethal discrimination to girls who are prevented from being born or disadvantaged from birth. This subject requires careful language: demographic imbalance is not evidence that every family or community holds identical beliefs, and India has also made legal, educational and social progress. Nevertheless, UNFPA has repeatedly identified gender-biased sex selection and son preference as continuing concerns. The preference is sustained by interlocking ideas about inheritance, lineage, care in old age, dowry and the economic value assigned to sons and daughters.

Bangladesh, Pakistan, Nepal and other South Asian societies present related but distinct combinations of intimate-partner violence, child marriage, acid attacks, honour-related killings, labour exploitation and barriers to justice. The category “South Asia” should help us notice regional patterns without pretending that a woman’s life in rural Pakistan is interchangeable with one in Mumbai, Dhaka or Kathmandu.

Afghanistan, Iran and state-enforced gender inequality

Afghanistan is an exceptional contemporary example of danger embedded directly in the state. Since returning to power in 2021, the Taliban has progressively excluded women and girls from secondary and higher education, restricted employment, constrained movement, imposed sweeping rules on dress and conduct, and removed women from large areas of public life. The cumulative effect has been widely described by Afghan women and international experts as gender apartheid: not merely the presence of discriminatory laws, but the organised subordination and segregation of women as a class.

This form of danger cannot be measured by homicide rates alone. A state can destroy a woman’s autonomy, education, livelihood, health and public existence without killing her. Restrictions also create indirect physical risks by deepening poverty, limiting access to healthcare, obstructing women’s ability to seek assistance and forcing dependence on male relatives. UN Women’s reporting on Afghanistan has documented the scale of women’s exclusion and the widening gap between Afghan women and men across health, education, financial inclusion and decision-making.

Iran presents a different political and social system and should not be treated as interchangeable with Taliban-ruled Afghanistan. Compulsory veiling, discriminatory family law, state violence against protesters and the policing of women’s bodies remain serious concerns, while Iranian women continue to organise, resist and participate in society in ways an external stereotype often fails to capture. Other parts of West and Central Asia present their own mixtures of family law inequality, honour-related violence, restrictions on autonomy and uneven legal protection.

These realities deserve scrutiny. The analytical mistake occurs when highly visible state or religious restrictions are treated as the complete definition of danger to women. State-enforced inequality and interpersonal femicide are distinct dimensions. A woman living under a restrictive legal regime is not automatically more likely to be murdered than a woman living in a country with formal legal equality. Conversely, the presence of equal-rights legislation does not prove that women are physically safe.

This distinction creates a far stronger criticism of selective women’s-rights discourse than denial ever could. The Taliban’s oppression of Afghan women is real. So is epidemic partner violence in societies that describe themselves as liberal, modern and equal. Women do not benefit when one truth is used to invalidate the other.

Conflict zones: when women’s bodies become part of the battlefield

War creates a further map of danger that ordinary national comparisons cannot capture. Conflict does not invent gender inequality from nothing; it intensifies existing hierarchies while destroying the systems women might otherwise use for protection. Police collapse, courts cease functioning, health facilities are attacked, displacement separates families, armed groups acquire power over food and movement, and sexual violence can become a method of terror, punishment, forced displacement or ethnic persecution.

In the eastern Democratic Republic of the Congo, decades of conflict have exposed women and girls to rape, sexual slavery, abduction, forced marriage, exploitation and repeated displacement. The scale is difficult to quantify precisely because survivors face insecurity, stigma, destroyed services and barriers to reporting. UN Women’s 2025 briefing on women and girls in the DRC described escalating sexual violence amid renewed conflict and warned that women’s protection and health services were being overwhelmed.

Sudan’s war has likewise produced reports of widespread conflict-related sexual violence, abduction, forced marriage, trafficking and attacks on displaced women and girls. The violence occurs alongside hunger, mass displacement, collapsed healthcare and attacks on civilian communities. A woman who dies from injuries following rape, during a forced pregnancy, while attempting to flee, or because reproductive healthcare has been destroyed may never enter a dataset labelled femicide, even when gender shaped every stage of her death.

This is one of the hardest conceptual problems in mapping femicide. When women are deliberately targeted with sexual violence during war, murdered after rape, killed because of their association with an ethnic or political group, or subjected to deprivation that falls differently on women and girls, the boundary between femicide, war crime, crime against humanity and conflict mortality becomes unstable. The legal categories matter for accountability, but the separation can fragment our understanding of what killed her.

Conflict also introduces a warning against treating national averages as stable truths. A country may contain relatively secure areas and regions in which women face catastrophic danger. Displaced women, women from persecuted ethnic communities, women with disabilities and those living without male protection can experience risks bearing little resemblance to the experience of women elsewhere within the same border.

The wealthy West is not the control group

Europe, Canada, the United States, Australia and New Zealand generally benefit from stronger institutions, greater formal legal equality and lower regional rates of intimate-partner and family-member femicide than many other parts of the world. That matters. It would be intellectually dishonest to imply that every society is equally dangerous to women, just as it would be dishonest to mistake a lower average for safety.

The EU gender-based violence survey, covering all 27 Member States, confirms that physical violence, threats, sexual violence, intimate-partner abuse and workplace sexual harassment remain widespread across Europe. Europe recorded the lowest estimated regional rate of intimate-partner and family-member femicide in 2024, at approximately 0.5 victims per 100,000 women and girls, but a large proportion of female homicide victims in the region were still killed by intimate partners. Lower prevalence is not the absence of a gendered pattern.

Britain offers a particularly clear example of the gap between formal equality and private danger. Women can vote, work, study, travel and own property, but coercive control, stalking, sexual violence and post-separation abuse continue to end women’s lives. A woman does not become safe because the society around her recognises her legal personhood if police fail to act on repeated warnings, courts underestimate stalking, an abusive partner can locate her, or professionals continue to ask why she did not simply leave.

National averages also conceal intersectional danger. Canada’s National Inquiry into Missing and Murdered Indigenous Women and Girls concluded that persistent and deliberate human-rights violations, colonial policies, state failures and social marginalisation had produced a national crisis. Indigenous women and girls face risks that cannot be inferred from Canada’s overall position in a global gender index. Similar disparities affect Indigenous women in the United States and Australia, Black women in multiple jurisdictions, migrant women whose immigration status is controlled by an abuser, disabled women dependent on carers, transgender women and women in prostitution or other criminalised economies.

This is why the wealthy West cannot be treated as the neutral control group against which the rest of the world’s misogyny is measured. Wealth, democratic institutions and liberal law can reduce danger, and often do. They can also coexist with entrenched male entitlement, racialised abandonment, institutional disbelief and a cultural tendency to frame domestic killings as unpredictable relationship tragedies rather than the fatal endpoint of known abuse.

The women missing from every femicide map

Every global map is incomplete because entire categories of women’s deaths remain statistically unstable or invisible. Women who disappear may never be counted as homicide victims. Unidentified bodies cannot easily be connected to relationships, reports of stalking or histories of exploitation. Women killed by traffickers, clients, pimps or organised criminal groups may be recorded under general homicide. A woman murdered after rejecting a man’s advances may appear as the victim of a stranger crime without the killer’s sexual entitlement being coded as a gender-related motive.

So-called honour killings and dowry deaths may occupy special legal categories in one jurisdiction and generic homicide categories in another. Women accused of witchcraft or sorcery can be murdered within communal violence that is gendered but not recorded as femicide. Family annihilation can be reported primarily through the perpetrator’s psychological state, obscuring a history of coercive control and the fact that a man killed his partner and children as a final exercise of ownership. Deaths connected to unsafe abortion, reproductive coercion, trafficking or sexual exploitation can sit outside homicide statistics entirely even when gendered power was fundamental to the fatal outcome.

Then there is abuse-related suicide. A woman may live for years under coercive control, threats, isolation, degradation, financial abuse, sexual violence and physical assault before dying by suicide. Conventional crime statistics record a self-inflicted death; the abusive partner disappears from the homicide data. The causal and legal questions are complex, and not every suicide by a victim-survivor can be attributed to an abuser. However, the fact that India separately records thousands of cases of “abetment to suicide of women” demonstrates that legal systems already recognise circumstances in which another person’s conduct can contribute directly to a death classified as suicide.

Any serious account of femicide must therefore acknowledge both the value and the limits of classification. Expanding the concept until every preventable death of a woman becomes femicide would make it analytically meaningless. Restricting it only to killings already proved and prosecuted as gender-motivated, however, would reproduce the failures of the systems being examined. The responsible position lies between: use clear definitions, distinguish categories, preserve uncertainty and remain alert to the deaths that existing categories are structurally designed to miss.

The most dangerous man is often not a stranger

After travelling across continents in search of the most dangerous place to be a woman, the evidence repeatedly returns us to a far smaller geography: the home, the relationship and the period surrounding separation. The global estimate that 60 per cent of intentionally killed women and girls in 2024 died at the hands of partners or family members forces us to reconsider the safety advice given to women from childhood. Women are taught to fear dark streets, isolated paths, strangers, taxis and nightclubs. Those dangers are real, but they can obscure the fact that the man most capable of locating, monitoring and killing a woman may already know where she sleeps.

Intimate-partner femicide is frequently preceded by a history of coercive control, stalking, threats, escalating assault, sexual violence, strangulation, weapon access, obsessive jealousy or threats to kill. Separation can increase danger because leaving challenges the perpetrator’s control. The relationship may be over for the woman while remaining, in his mind, a system of ownership he is entitled to enforce.

When the perpetrator cannot reach her directly, he may redirect his aggression towards her children, relatives, friends, new partner, animals or property. These attacks are sometimes treated as secondary or unrelated incidents. In reality, they can be communications to the woman: demonstrations that distance has not ended his power and that anything she loves can be reached. The pattern matters because risk does not begin with an assault on her body and does not end when she changes address.

Across the cases examined throughout my femicide research, three motives recur with particular force: control, jealousy or possession, and material or financial interest. They are not universal explanations, and the categories frequently overlap. “Jealousy,” especially, can sanitise entitlement by making murder sound like an excess of love. A man who kills a woman because she left, rejected him or formed a new relationship is not expressing attachment; he is enforcing a belief that her autonomy is an offence against him.

This is also why the familiar question “Why didn’t she leave?” is dangerously inadequate. It assumes that departure ends abuse, that resources are available, that children can be moved safely, that police will protect her, that the perpetrator will respect legal boundaries and that he does not regard separation as the ultimate provocation. The more useful question is why he believed he had the right to control her—and why the institutions around her failed to recognise what that belief could become.

Who gets to decide what women’s oppression looks like?

International concern about women’s rights is never politically neutral. Governments, media organisations and campaigners make choices about which women become symbols, which laws become proof of barbarism, which deaths become global news and which forms of violence remain domestic background noise. Those choices are influenced by race, religion, colonial history, foreign policy and the availability of familiar images.

A morality law, compulsory garment or ban on girls’ education offers a visible symbol of oppression and a clearly identifiable state actor. The violence is legible to an external audience. Coercive control inside a private home is less visible. So is a woman disappearing along a migration route, an Indigenous woman dismissed as “high risk,” a Pacific Islander living hours from the nearest service, a bride subjected to dowry demands, or a woman whose suicide follows years of terror. These harms are no less real because they are harder to turn into a political emblem.

Western states have a long history of invoking the condition of women elsewhere to legitimise intervention, occupation or claims of civilisational superiority. That history gives us good reason to scrutinise selective outrage, but it must not lead us into denying the words of women living under repressive regimes. Afghan and Iranian women do not become inventions of Western propaganda because Western governments instrumentalise their suffering. The ethical response is to listen to them while applying the same seriousness to violence against women everywhere else.

The standard must therefore be consistent. If restricting a woman’s movement is recognised as an exercise of patriarchal power when a state does it, coercive control must also be recognised when a husband does it. If forcing a woman to cover is an assault on autonomy, forcing her to uncover, sexualise herself or submit to commercial exploitation cannot be repackaged automatically as liberation. If preventing girls from studying is understood as structural violence, so should the conditions that push girls out of school through child marriage, conflict, poverty or sexual danger. If the murder of a woman by a morality police shocks the world, the murder of a woman by a former partner who repeatedly threatened her should not be reduced to a domestic dispute that “went too far.”

The international imagination of the oppressed woman is selective because some forms of oppression confirm existing political stories and others expose the societies telling them. A genuinely global women’s-rights analysis has to withstand both temptations: the temptation to demonise particular cultures as uniquely misogynistic and the temptation to excuse severe abuses because criticism has sometimes been weaponised by imperial power.

So where are the most dangerous places in the world to be a woman?

There is no single honest answer, because there is no single form of danger. If the measure is intimate-partner and family-member femicide, Africa currently has the highest estimated regional rate. If the measure is lifetime intimate-partner violence, Oceania excluding Australia and New Zealand, sub-Saharan Africa and Southern Asia emerge particularly strongly. If the measure is legally or statistically identified femicide and feminicide, Latin America provides some of the world’s most developed recognition systems and records thousands of cases every year. If the measure is systematic exclusion from education, employment, movement and public life, Taliban-ruled Afghanistan is exceptional. If the measure is conflict-related sexual violence, Sudan, eastern DRC and other war zones demand immediate attention. If the measure is the intersection of colonialism, racism and gendered disappearance, Indigenous women can face levels of danger that bear little resemblance to the national average around them.

None of these conclusions cancels the others. They show why a serious global analysis cannot crown one country the winner of a contest no woman would wish her country to enter. The most dangerous countries for women are dangerous in different ways, and the danger experienced by women within the same country is divided by class, race, ethnicity, disability, sexuality, migration status, location and relationship circumstances.

The women found around Ekurhuleni were the reason I sat down to write this article, but they are not the whole article. Their deaths opened a door onto a larger question: how many different ways have human societies created for women to be unsafe, and why are we so much more willing to recognise some of them than others?

The final lesson of the global evidence is that geography is both essential and insufficient. Where a woman lives shapes the laws surrounding her, the services available to her, the prevalence of violence, the weapons accessible to perpetrators and the likelihood that the state will intervene. Yet for tens of thousands of women killed each year, the decisive geography is relational. The most dangerous place is somewhere she should have been safest, and the most dangerous person is someone she already knows.

Femicide is the final act, not the beginning of the story. Before the death there may have been stalking, strangulation, threats, sexual violence, coercive control, obsessive monitoring, weapon access, economic abuse, attacks on children or animals, and a refusal to accept separation. None of those behaviours guarantees a future homicide, but survival must never be mistaken for proof that the danger was insignificant.

The purpose of mapping femicide is not to become better at explaining women’s deaths afterwards. It is to recognise the patterns while women are still alive.

Take action: make women’s danger visible before it becomes fatal

Learn to recognise coercive control, stalking, strangulation, threats to kill, escalating violence, weapon access and post-separation abuse as potential indicators of serious danger. If a woman tells you she is afraid of a partner or former partner, treat her fear as evidence rather than hysteria. Do not pressure her to leave without considering safety planning, because separation may increase risk and a perpetrator may escalate when he believes control is being lost.

Support specialist organisations led by women in the communities they serve, including domestic-abuse services, women’s shelters, rape-crisis organisations, Indigenous women’s organisations and groups documenting femicide. In conflict settings, prioritise women-led local organisations that are already providing protection, healthcare, legal support and safe spaces. The Women’s Peace and Humanitarian Fund is one route for supporting women-led responses in crisis-affected countries.

Challenge media reporting that treats each murdered woman as an isolated tragedy, romanticises possessiveness as jealousy, centres the perpetrator’s distress, or interrogates the victim’s behaviour more aggressively than the warning signs preceding her death. Ask whether previous police reports, stalking, separation, threats, strangulation or coercive control were present, while respecting the privacy and dignity of the woman and her family.

Finally, demand better data. Governments should implement the UN statistical framework, link homicide records with histories of domestic abuse and stalking, record victim–perpetrator relationships, investigate missing women, and publish findings in forms that allow patterns to be recognised. Counting women properly will not save them on its own, but a society cannot prevent a pattern it has designed its records not to see.

If you are in immediate danger in the United Kingdom, call 999. If you cannot speak on a mobile call, press 55 when prompted. The National Domestic Abuse Helpline, run by Refuge, is available free on 0808 2000 247, 24 hours a day. Women’s Aid also provides information, safety-planning guidance and links to local services.

---

Frequently asked questions

---

Continue Exploring Global Femicide

If you enjoyed this article, you might also like to read Safer Because of You: The Positive Alternative to the Manosphere and Thailand Sex Tourism: How War, Tourism, and Foreign Male Demand Built Pattaya’s Exploitation Economy

Please also explore the Mapping Femicide Media List I have curated especially for this post, which offers a series of books, films, and documentaries that explore different aspects of global femicide.

Sources and further reading

• UN Women and UNODC: Femicides in 2024—Global estimates of intimate partner/family member femicides
• UN Women and UNODC: Statistical framework for measuring the gender-related killing of women and girls
• World Health Organization: Violence against women—global facts and prevalence estimates
• ECLAC: Femicide and feminicide in Latin America and the Caribbean
• South African Medical Research Council: Twenty years of national femicide research
• UN Women Data Hub: Papua New Guinea country data
• Government of India: Crime cases against women, 2013–2022
• UNFPA India: Gender-biased sex selection
• UN Women: State of women’s rights in Afghanistan
• UN Women: The situation of women and girls in the Democratic Republic of the Congo
• Eurostat, FRA and EIGE: EU gender-based violence survey—key results
• National Inquiry into Missing and Murdered Indigenous Women and Girls: Final Report
• Reuters: South African police find a ninth woman’s body near Johannesburg
• Reuters: South Africans gather to mourn nine women and demand change

---

Thank You for Reading!

I hope you enjoyed this post and found it insightful. If you did, feel free to subscribe to receive updates about future posts via email, leave a comment below, or share it with your friends and followers. Your feedback and engagement mean a lot to me, and it helps keep this community growing.

If you’re interested in diving deeper into topics like this, don’t forget to follow me on Instagram @jessielouisevernon, though my accounts have been shut down before (like my old @doncvoilaquoi on Instagram), so keep an eye out for updates.