Reading time: 3 minutes, 27 seconds (860 words)
When assessing risk in women, mental health professionals should remain aware of gender’s influence on risk factors — both their presence and relevance to the overall picture of risk. They should keep in mind that when women commit crime and violence, these actions are often related to their unique experiences and social roles.
Women and girls differ from men and boys in many aspects of criminal and violent behaviour, said Dr. Stephanie Penney in a presentation for the International Forensic Psychiatry Lecture Series. Their development before committing violence can differ. There are differences in who their victims are, the pathways they take to justice involvement, and the motivations for crime and violence.
As a result, risk assessment measures may not reflect the risk factors for women. In assessing women, mental health professionals should avoid using some measures and carefully consider results of others to improve predictions of future violence.
Gender Differences
This population is much less likely to commit crimes in general — and specifically violent crimes — than men and boys. This difference is seen over time and across countries and cultures. When the frequency of crime and violence is examined in people with complex and serious forms of mental disorder such as schizophrenia, women are found to be as likely to engage in violence.
Women and girls who become involved with the criminal justice system have higher rates of mental health needs and mental illness. They also experience higher rates of victimization and adverse childhood events. In fact, one study found that such experiences are almost universal in women involved in the criminal justice system. Such experiences may also have a stronger impact on women’s and girls’ psychological and social development than similar experiences have for men and boys. Furthermore, a widely held view of two different pathways to criminal behaviour — one limited to adolescence, the other starting early in life and becoming lifelong — may not apply to women.
In Canada, women account for about 15 per cent of forensic mental health patients. That proportion that has remained stable across time and different jurisdictions. In forensic populations, research suggests that this population has higher rates of convictions for homicide, attempted homicide, and arson than men. They are also found to be more likely to commit violence against their children and partners than against strangers.
A large study of women in forensic mental health care found that they have higher rates of mood and personality disorders, post-traumatic stress disorder, suicidality, and self-injury. Women’s onset of mental illness is later in life. As a result, they may have been married, had children, and had jobs. They also spend less time in forensic care. They are less likely to have previous criminal convictions, and they have lower rates of substance use.
It is unclear whether women in the forensic mental health system differ from men in their rate of re-offending after being discharged. Some studies have found no difference. Others have found women have lower rates of re-offending, and one study found they had a higher rate.
Risk Assessment in Women
Mental health professionals should be aware of several issues that arise when assessing violence risk in women. Professionals tend to underestimate risk in this population, according to several studies. Many research studies of risk assessment measures have included too few women in their samples to reach statistical conclusions with confidence.
Dr. Penney recommends avoiding the use of actuarial risk assessment measures when assessing risk in women. For example, the Violence Risk Assessment Guide–Revised can lead to incorrectly high predictions of risk in women. This measure was developed and standardized in men, who have a higher base rate of violence. When it is used in a population where the base rate of violence is lower, it incorrectly estimates a higher risk level than what is actually observed in practice.
The Historical-Clinical-Risk Management 20 Version 2 or 3 (HCR-20V2/V3) is considered appropriate in assessing risk in women, and its use in this population is supported by several research studies. The dynamic clinical and risk management factors appearing on this measure, as well as the item assessing trauma and victimization history, are often particularly relevant for this population.
A risk assessment guideline for women, the Female Additional Manual, adds several items to the HCR-20V3. These include historical items relevant to women and three risk ratings specific to women. In studies of this manual, some found it did not add to the predictive value of HCR-20V3. One study found that it was moderately valid for predicting re-offending, but its validity in predicting repeated violence in women was low.
Dr. Penney concluded that existing measures are still essential for assessing risk in a structured and transparent manner. However, she advised that mental health professionals acknowledge that their accuracy in predicting violence in women is moderate and often less consistent than in men. They should interpret results with caution.
For more details, watch the recording of the April 2, 2026, lecture, “Research and Best Practices in Risk Assessment for Women” on our website.
Written by Carolyn Brown and approved by Dr. Stephanie Penney.
