Filicide is the killing of a child by a parent, and mothers who kill their children are a rare but distinct subgroup. Most cases fall into two broad patterns: a psychotic or acutely mentally ill mother acting on delusions, and a mother in an extreme psychosocial crisis, such as a violent partner or a murder-suicide dynamic. Neither pattern is explained by "evil" or a single breakdown. Both reflect measurable failures of mental-health care, intimate-partner violence screening, and child-protective referral systems that research has mapped in detail.
The question most people actually carry is simpler: how does a mother end up killing her own children, and could anyone have seen it coming? The scientific literature on maternal filicide gives partial answers. It identifies recurrent mechanisms, quantifies how often each appears, and shows where clinicians and family members have missed signals. It also shows that prediction at the individual level remains weak, which is why the strongest evidence sits in prevention systems rather than in profiling.
Mechanism: How Maternal Filicide Happens
Filicide is the killing of a child by a biological, adoptive, or step-parent. In the research literature, maternal filicide is usually divided into two broad etiologic types. The first is the psychotic type: the mother is experiencing a severe mental illness, most often schizophrenia-spectrum or psychotic bipolar disorder, and the killing is driven by delusions or command hallucinations. She may believe the child is possessed, doomed, or being harmed, and killing becomes a protective act inside her altered reality.
The second is the "normal" or psychosocial type, where no major psychosis is present. Instead, the killing emerges from accumulated hopelessness, rage, or coercion. This type includes fatal child abuse, neonaticide in the first 24 hours of life, and murder-suicide in the context of a violent partner or impending family collapse.
The depth here is in how the psychotic pathway works. A delusional mother often presents to clinicians with what looks like mild distress, not overt horror. Because the delusion feels true to her, the urge to "save" the child appears rational and therefore maskable. The result is a paradox that recurs in case reviews: the most dangerous mothers are often the ones whose illness is most contained, because the logic of the delusion is intact enough to hide.
A second mechanism is psychological entrapment. Research on filicide-suicide describes a cognitive state in which the mother concludes that death is the only escape and that the children must die with her, either to spare them suffering or to keep them "with her." This is not a snap decision. Reviews of maternal filicide-suicide find it is more often planned than the public assumes.
Quantified Context: Who, When, and How Often
Filicide is rare in absolute terms but not negligible. A frequently cited review of the international literature estimated that about 15 percent of filicide offenders are mothers, though estimates vary widely by country and by how the deaths are classified. Across high-income countries, filicide rates cluster around 1 to 8 deaths per 100,000 children per year, depending on the source. The Global Burden of Disease work on child homicide gives a broader comparable figure: globally, child homicide rates in recent estimates have hovered near 1.7 per 100,000 children under age 18.
What drives a media image is a specific subtype: maternal filicide involving an older child, usually older than 5 years. Yet the data show the opposite profile is more common. The World Health Organization notes that the risk of child homicide is highest in infancy. One review of maternal filicide in high-income countries found that infants under 1 year accounted for a large share of maternal filicide victims. Neonaticide, the killing of a baby in the first 24 hours of life, is almost exclusively maternal.
This inversion is the first counterintuitive detail worth keeping. A typical mother in the research literature is more likely to be young, unpartnered, and killing a newborn than her older children.
Timing matters. Reviews find that more than half of maternal filicide victims in some series are under 1 year. Neonaticide itself usually happens within the first 24 hours of delivery, and a large fraction of neonaticide cases involve concealed or denied pregnancy. In the case of older victims, the most common time frame is the months after a separation, a custody dispute, or a marital crisis.
Weapon patterns track gender and access. A 2015 systematic review of maternal filicide in high-income countries found that mothers most often killed by suffocation, drowning, or gassing, and murder-suicide was a recurrent feature. That pattern differs from paternal filicide, which more often involves a violent strike or a weapon, and it explains why many maternal cases look like a "quiet" act in the home.
Timeline: The Time Course of Maternal Filicide
| Phase | Typical time frame | What research finds | Window for intervention |
|---|---|---|---|
| Denied or concealed pregnancy | Weeks to months | Often no prenatal care; fear of family or partner discovery | School, employer, or clinic detection; crisis pregnancy services |
| Perinatal crisis | Third trimester to 48 hours post-delivery | Psychosis onset, extreme exhaustion, no antenatal psychiatric contact | Postpartum emergency screening; maternity ward mental-health referral |
| Early infancy | First 6 months | Sustained depression, detachment from the infant, infant crying as trigger | Home-visiting programs; pediatric well-child checks |
| Acute escalation | Days to weeks | Psychotic ideas, command hallucinations, obsessive protection ("I must save them") | Emergency assessment; same-day psychiatric admission |
| Separation or custody crisis | Weeks after a breakup | Entrapment thinking, murder-suicide planning | Domestic violence and family court screening |
| Prior attempt or ideation | Hours to weeks | A prior attempt or threats to the children are a recorded signal | Immediate safety plan and child welfare check |
| Post-event | Within hours | Injury patterns, drowning, or suffocation found on autopsy | Forensic and child-protection follow-up |
This table is a schematic synthesis of the case-review literature, not a set of defined clinical thresholds. While many of these windows exist, they overlap.
Edge Cases That Change the Picture
The most common error is to treat the phrase "woman who killed her 3 kids" as describing one kind of case. The research base argues against that. The subtypes differ in mechanism, victim age, mental state, and prevention point, and lumping them together produces poor prevention strategy.
Neonaticide and maternal filicide of older children should be studied separately. Neonaticide is characterized by denial or concealment of pregnancy, a lack of antenatal care, and a killing within 24 hours of birth. In those cases, the mother often has no prior psychiatric diagnosis. By contrast, filicide of older children is much more likely to involve serious mental illness, prior warning signs, and maternal suicide attempts.
Mental illness is the single most cited risk factor in older victims. Reviews consistently report that psychotic illness is present in a substantial minority, with estimates varying substantially by sample. Studies that used structured diagnostic interviews often found a diagnosed mental disorder in the majority of maternal filicide cases. The variation in these numbers is not noise. It reflects how the cases were sampled and how strictly psychosis was defined.
Substance use, young maternal age, and intimate-partner violence are also documented risk factors. In cases where domestic violence is present, the child's death is sometimes an extension of a partner-directed attack or of a murder-suicide that included the children.
A professional caveat that appears repeatedly in forensic psychiatry: case reviews of filicide are drawn from police, hospital, and court records, which means they overrepresent the cases that ended in formal investigation. Cases that looked accidental, or where the mother did not survive, are underrepresented. The absence of a risk factor in a review does not mean it was absent in the case.
Misconceptions to Correct
The first misconception is that filicidal mothers are obviously deranged. The case literature shows the opposite far more often. Mothers with delusional beliefs are often composed and formulaic in their presentation to clinicians and family. That containment is precisely why family members describe the act as sudden.
The second misconception is that a woman who kills her children must have hated them. In the psychotic subtype, the enduring motive is often hyperprotective. The mother believes she is saving the child from a worse fate. This does not reduce the act to something other than terrible. It does change the physician's question, which becomes: does this mother believe the child is in danger, and does she believe that danger is real?
The third misconception is that family members and clinicians miss cases because people "don't ask." In practice, the gap is often structural: understaffed maternity and perinatal services, no scheduled postpartum psychiatric review, and no pathway connecting a domestic violence disclosure to a child safety check. Closing the gap requires scheduled contact points, not a better questionnaire alone.
The fourth is that a murder-suicide in which the mother survives is always a failed suicide. Data show that in many filicide-suicide cases, the mother intended both and simply survived. Treating her surviving as evidence of a lesser intent distorts both the legal and clinical picture.
Practical Actions: What Actually Helps
Prevention is the only place the evidence supports specific steps, and the steps are mostly scheduling and screening, not intuition.
Treat a mother's stated fear that she might harm her children as an emergency psychiatric presentation. In cases with psychotic content, this has sometimes been the only warning. A same-day evaluation and referral is the practical threshold. Screen for perinatal psychosis with standardized tools at the postpartum visit. Psychotic symptoms that begin within the first weeks after delivery deserve the same urgency as a fever in a newborn. The highest-frequency window is the first three to six months. Link a domestic-violence disclosure to a child-safety check. The entrapment pathway runs through partner violence. A disclosure that ends at a referral without a plan for the child is an incomplete intervention. In custody disputes, flag threats to harm the children as high-priority safety-plan triggers, not as manipulation. A custody evaluation that does not screen for filicidal statements is missing a known signal. Never promise a neat explanation of filicide. Individual prediction of maternal filicide remains weak. The evidence supports risk-informed systems, not risk-score profiling.
Limitations and Warnings
The literature on maternal filicide has real limits. Samples are small, drawn mostly from high-income countries, and skewed toward cases that resulted in prosecution or hospital admission. Base rates are so low that individual prediction cannot be supported. The subtypes are overlapping and the terminology is inconsistent across studies. Finally, research on this topic is drawn from case reviews rather than prospective studies, because prospective research on child homicide is ethically impossible.
Do not use this article as a diagnostic or risk-profiling tool. It is general educational information, not a diagnosis, treatment plan, or legal opinion, and it cannot predict whether any particular person is dangerous. If you are worried that someone may be a danger to themselves or a child, contact local emergency services or a crisis line and a qualified mental-health professional immediately.