Feeling unsafe? Find support services   emergency? call 000


Our research

Violence against women and their children affects everybody. It impacts on the health, wellbeing and safety of a significant proportion of Australians throughout all states and territories and places an enormous burden on the nation’s economy across family and community services, health and hospitals, income-support and criminal justice systems.


News and events

ANROWS hosts events as part of its knowledge transfer and exchange work, including public lectures, workshops and research launches. Details of upcoming ANROWS activities and news are available from the list on the right.



ANROWS was established by the Commonwealth and all state and territory governments of Australia to produce, disseminate and assist in applying evidence for policy and practice addressing violence against women and their children.



To support the take-up of evidence, ANROWS offers a range of resources developed from research to support practitioners and policy-makers in delivering evidence-based interventions.

National Risk Assessment Principles for domestic and family violence

Quick reference guide for practitioners

Read the PDF version of the report

Principle 1
Survivors’ safety is the core priority of all risk assessment frameworks and tools.

The safety and wellbeing of adult and child survivors of domestic and family violence (DFV) is the first priority of any response. Risk must be identified, comprehensively assessed and appropriately responded to by holding the perpetrator responsible and accountable for their behaviour and actions.

Principle 2
A perpetrator’s current and past actions and behaviours bear significant weight in determining risk.

While the safety of adult and child survivors of DFV is prioritised, workers must also reorient risk assessment and safety management processes onto the behaviour of perpetrators, rather than focusing solely on the protective strategies of survivors.

Principle 3
A survivor’s knowledge of their own risk is central to any risk assessment.

A survivor’s assessment of their own risk should be considered one of the primary elements of any risk assessment, as it provides intimate knowledge of their lived experience of violence and patterns of coercive control.

Service providers need to approach risk assessment and safety management with adult and child survivors through a collaborative process which respects and builds on the survivor’s own assessment of their safety, as well as drawing on other sources of information. These sources may include: the use of a well tested actuarial risk assessment tool; professional judgement and practice wisdom drawn from workers’ specialist knowledge of domestic and family violence; and information gathered from other organisations.

Principle 4
Heightened risk and diverse needs of particular cohorts are taken into account in risk assessment and safety management.

Some members of diverse communities are more vulnerable to DFV, experience violence more frequently and with more severity than others and face a range of specific barriers to safety. An understanding of the effect of the intersections of gender, ethnicity, sexuality, disability, culture, mental health issues, citizenship, age, economic status, geographical isolation and other identity-based and situational factors are critical when undertaking risk assessment and managing safety.

Principle 5
Risk assessment tools and safety management strategies for Aboriginal and Torres Strait Islander peoples are community-led, culturally safe and acknowledge the significant impact of intergenerational trauma on communities and families.

It is important to work with extended families and communities in responding to Aboriginal and Torres Strait Islander family violence. Workers need to respond to the whole family rather than to individuals. Healing for adult and child survivors, as well as for the perpetrators of violence, is key to all responses, including risk assessment and management.

Community-driven, trauma-informed approaches to family violence, which prioritise cultural healing and are based on the understanding that culture is a key protective factor supporting Aboriginal and Torres Strait Islander families to live free from violence, are critical.

Principle 6
To ensure survivors’ safety, an integrated, systemic response to risk assessment and management, whereby all relevant agencies work together, is critical.

Working collaboratively across agencies is fundamental to improving the safety and wellbeing of adult and child survivors. This can be best achieved through an integrated, systemic response that ensures all relevant agencies work together on risk assessment and risk management processes in partnership with the survivor. Effective leadership and governance arrangements which support collaboration and partnerships are essential for collaborative service delivery.

Principle 7
Risk assessment and safety management work as part of a continuum of service delivery.

Risk assessment should always form part of a safety management approach which moves with the adult or child survivor on their journey away from violence. The development of a continuum of service responses which addresses survivor safety, perpetrators taking responsibility for their violence and aspects of prevention and healing is critical. As risk factors change over time, ongoing risk assessment and management along the service continuum also changes.

Principle 8
Intimate partner sexual violence must be specifically considered in all risk assessment processes.

Intimate partner sexual violence (IPSV) is a uniquely dangerous form of DFV which must be specifically considered in all risk assessment and safety management processes and practices. Survivors’ who are sexually abused by their partners are at a much higher risk of being killed, particularly if they are also being physically assaulted, and IPSV is a significant indicator of escalating frequency and severity of domestic and family violence.

More so than other factors, IPSV is under-reported and often not disclosed. Training on IPSV for all workers conducting DFV risk assessment is essential. Training should include:

  • details on the myths and dynamics of sexual violence within relationships;
  • guidance on “how to ask” sensitively and building trust;
  • the specific effects and health consequences of IPSV;
  • how best to manage victim survivors’ safety;
  • cultural considerations; and
  • legal options and evidence requirements.

Principle 9
All risk assessment tools and frameworks are built from evidence-based risk factors.

The factors critical to developing a shared understanding of risk and safety include:

  • Evidence-based risk factors: variables which assist in assessing the likelihood that violence will be repeated or escalate and responding appropriately to that violence.
  • Conditions of vulnerability: identity-based and situational factors which may indicate heightened vulnerability to violence, and which may intersect with other factors to compound the risks and effects of violence.
  • Protective factors: characteristics which mitigate or eliminate risk, or which reduce conditions
    of vulnerability.
  • Determining a risk threshold: identification of “risk” or “high-risk” through a thorough assessment, so that the allocation of support and treatment interventions address the specific needs of individual survivors and perpetrators.

Specific evidence-based risk factors and their impact on determining risk thresholds are outlined in the following table: High-risk factors for domestic and family violence.

High-risk factors for domestic and family violence

National Risk Assessment Principles Quick reference guide for practitioners

There are many factors which contribute to the risk of domestic and family violence (DFV). However, findings from empirical studies, academic and practice-based literature, and reports produced by international and Australian domestic violence death review committees and Coroner’s Courts indicate that some risk factors are associated with a higher likelihood of violence reoccurring, serious injury, or death, in the context of intimate partner violence by men against women.[1] The relationship between these factors and risk of reassault or lethality are not always straightforward, and no one factor can be considered singularly “causal”. Importantly, there are diverse forms of DFV that do not necessarily involve risk of physical violence or lethality, but which can have a devastating impact on victims’ lives. While there is significant evidence that the below risk factors indicate high risk of serious harm or death when mediated by other risk factors or an individual’s situation, all of these factors are salient in any case of DFV and should be responded to appropriately and proportionately, whether or not there is a clear intent of homicide.

Lethality/High-risk factors

Factor Key facts
History of family and
domestic violence
  • The most consistently identified risk factor for intimate partner lethality and risk of reassault is the previous history of violence by the perpetrator against the victim.
  • In their 11-city study in the United States (US), Campbell et al. (2003) found that 72 percent of intimate partner femicides were preceded by physical violence by the male perpetrator. When there was an escalation in frequency or severity of physical violence over time, abused women were five times more likely to be killed.
  • Smith, Moracco, & Butts (1998) found that for 75 percent of homicides perpetrated by women, the relationship was characterised by a history of abuse by her male partner and the homicide was preceded by male-initiated violence.
  • Homicide is rarely a random act and often occurs after repeated patterns of physical and sexual abuse and psychologically coercive and controlling behaviours.
(actual or pending)
  • Women are most at risk of being killed or seriously harmed during and/or immediately after separation.
  • The NSW Domestic Violence Death Review Team recorded that two-thirds (65%) of female victims killed by a former intimate partner between 2000-2014, had ended their relationship within three months of the homicide.
  • Separation is particularly dangerous when the perpetrator has been highly controlling during the relationship and continues or escalates his violence following separation in an attempt to reassert control or punish the victim.
  • Children are also at heightened risk of harm during and post-separation.
Intimate partner
sexual violence
  • Intimate partner sexual violence (IPSV) is a uniquely dangerous form of exerting power and control due to its invasive attack on victims’ bodies and the severity of mental health, physical injury and gynaecological consequences.
  • Campbell et al. (2003) found that physically abused women who also experienced forced sexual activity or rape, were seven times more likely than other abused women to be killed and IPSV was the strongest indicator of escalating frequency and severity of violence, more so than stalking, strangulation and abuse during pregnancy.
  • The 2016 ABS Personal Safety Survey (PSS) found that since the age of 15, 5.1 percent (480,200) of Australian women have experienced sexual violence by a partner. Heenan (2004) found that Australian domestic violence workers believe that 90-100 percent of their female clients have experienced IPSV.
  • More than other factors, IPSV is under-reported by victims. Shame and stigma caused by commonly held assumptions that discussing sex or sexual assault within relationships is “taboo”, are significant barriers to seeking help for IPSV.
Non-lethal strangulation
(or choking)
  • Strangulation is one of the most lethal forms of intimate partner violence. When a victim is strangled, whether by choking or other means of obstructing blood vessels and/or airflow to the neck, they may lose consciousness within seconds and die within minutes.
  • Glass et al. (2008) found that women whose partner had tried to strangle or choke them were over seven times more likely than other abused women to be killed, whether by repeat strangulation or another violent act.
  • The seriousness of strangulation as an indicator of future lethality is often misidentified, or not responded to proportionately, as a consequence of the often minimal visibility of physical injury. However, many victims suffer internal injuries which may result in subsequent serious or fatal harm.
  • Most perpetrators do not strangle to kill but to show that they can kill. Non-lethal strangulation is a powerful method of exerting control over victims. Through credible threat of death, perpetrators coerce compliance.
  • Stalking behaviours (repeated, persistent and unwanted) including technology-facilitated surveillance, GPS tracking, interferences with property, persistent phoning/texting and contact against court order conditions, increases risk of male-perpetrated homicide.
  • The 2016 ABS PSS found that since the age of 15, one in six Australian women (17% or 1.6 million) have experienced at least one episode of stalking.
  • McFarlane et al. (1999) found that stalking was a factor in 85 percent of attempted femicides and for 76 percent of femicide victims.
  • The vast majority of perpetrators of stalking, and the most dangerous, are intimate partners of the victim, and not
    a stranger.
Threats to kill
  • Perpetrators who threaten to kill their partner or former partner, themselves or others including their children, are particularly dangerous. Threats of this nature are psychologically abusive.
  • Campbell et al. (2003) found that women whose partners threatened them with murder were 15 times more likely than other women experiencing abuse to be killed.
  • Humphreys (2007) found that actual attempts to kill are difficult to separate from serious physical and sexual
    abuse, and that as above, attempted strangulation is of particular concern given the prevalence of femicide
    through strangulation.
Perpetrator’s access to,
or use of weapons
  • Use of a weapon (any tool used by the perpetrator that could injure, kill or destroy property) indicates high risk, particularly if used in the most recent violent incident, as past behaviour strongly predicts future behaviour.
  • Campbell et al. (2003) found that women who are threatened or assaulted with a gun or other weapon, are 20 times more likely than other abused women to be killed. The severity of abuse-related harm is significantly heightened when weapons are involved.
Escalation (frequency and/or severity)
  • The escalation in frequency and severity of violence over time is linked to lethality and often occurs when there are shifts in other dynamic risk factors, such as the attempts by the victim to leave the relationship.
  • Campbell et al. (2003) found that when there is an escalation in either frequency or severity of physical violence over time, abused women are more than five times more likely to be killed.
  • Dwyer and Miller (2014) found that police investigations and family, criminal or civil court proceedings can trigger an escalation in the aggressive and violent behaviour of the perpetrator and heighten risk to the partner and children. Transition points such as this should be treated with great caution.
Coercive control
  • Reports from death review committees and Coroner’s Courts highlight the prevalence of patterns of coercive and controlling behaviours prior to male-perpetrated intimate partner homicide, including verbal and financial abuse, psychologically controlling acts and social isolation.
  • Elliott (2017) found through a synthesis of key empirical research, that coercive control is a gendered pattern of abuse, and is the primary strategy used to coerce and exercise control over female survivors by a current or former male partner. Understanding violence as coercive control, highlights that it is ongoing, cumulative, chronic and routine.
  • Coercive and controlling patterns of behaviours are particularly dangerous and can heighten the risk of lethality, in contexts where other high-risk factors are present, such as attempts by the victim to leave the relationship.
Pregnancy and new birth
  • Violence perpetrated against pregnant women by a partner is a significant indicator of future harm to the woman and child, and is the primary cause of death to mothers during pregnancy, both in Australia and internationally.
  • The 2016 ABS PSS found that nearly half (48% or 325,900) of women who have experienced violence by a previous partner and who were pregnant during that relationship, experienced violence from their partner while pregnant.
  • Humphreys (2007) highlights this violence as “double-intentioned”, where perpetrators may aim physical violence at their partner’s abdomen, genitals or breasts, so that abuse is both of the mother and child.
  • Women with a disability, women aged 18-24 years and Indigenous women are at particularly significant risk of experiencing severe violence from their partner during pregnancy.
  • Violence often begins when women are pregnant, and when previously occurring, it often escalates in frequency
    and severity.

Other Risk factors

Factor Key facts
Victim’s self-perception
of risk
  • A victim’s perception of their own risk of experiencing future violence is not sufficient by itself to accurately
    determine severity or incidence of violence. However, there is significant consensus across the literature that it is important to consider the victim’s own assessment as at a minimum, they can provide information relevant to their safety management.
Suicide threats
and attempts
  • Hart’s (1988) study found that the combination of attempts, threats or fantasies of suicide, availability of weapons, obsessiveness, perpetrator isolation and drug and alcohol consumption indicates severe or lethal future violence.
  • Threats of suicide, like most threats in the context of DFV, are a strategy used by perpetrators to exert control. The NSW Domestic Violence Death Review Team recorded that 24 percent of men who killed an intimate partner in NSW between 2000-2014 suicided following the murder.
Court orders and parenting proceedings
  • In their review of the Victorian Common Risk Assessment Framework (CRAF), McCulloch et al. (2016) found that from their experience, victims/survivors considered Family Law proceedings and intervention orders a critical and often overlooked indicator of DFV risk.
  • DFV is common and often escalates among separating parents. Perpetrators may use their joint parenting role or judicial options as a way of exercising control over their former partner.
Misuse of drugs or excessive
alcohol consumption
  • Alcohol and/or drug misuse and abuse are often exacerbating or moderating factors in predicting the dangerousness of a perpetrator, and may increase the severity of future violence.
  • Recent cessation of drug or alcohol use, particularly where addiction was present, can also exacerbate violent behaviour when the perpetrator is not actively involved in a recovery and rehabilitation process.
Isolation and barriers to help-seeking
  • Isolation, including limiting interactions with family, friends, social supports and community support programs is a control strategy used by some perpetrators and increases the risk of severe harm.
  • A victim is at increased risk of future violence if she has had no prior engagement with services and is presenting with DFV. A systematic review by Capaldi et al. (2012) found that social support and tangible help are protective against both perpetration and victimisation and that a lack of support is a significant risk factor for victims.
Abuse of pets and other animals
  • Cruelty and harm directed to pets and other animals can indicate risk of future or more severe violence and are often used as a control tactic by perpetrators.
  • Having to leave pets behind is a recognised barrier to victim-survivors leaving their violent partners.

Toivonen, C., & Backhouse, C. (2018). National Risk Assessment Principles for domestic and family violence: Quick reference guide for practitioners (ANROWS Insights 10/2018). Sydney, NSW: ANROWS


  1. Risk factors identified through empirical research have almost exclusively been identified using heterosexual, intimate partner samples, and their applicability to people in non-heterosexual LGBTQI relationships, or for violence occurring more broadly within families, remains unclear. In this resource, the terms “intimate partner violence” or “intimate partner lethality” have sometimes been used instead of “DFV” to accurately reflect the nature of the data source (such as the ABS Personal Safety Survey). Risk assessment practices and tools should be adapted in accordance with emerging knowledge and as further research determines how well the existing evidence-base applies to diverse relationships, families, communities and priority population groups.
Back to top