Victims/survivors of domestic violence often rely on social security payments for financial support when trying to leave a violent relationship and when trying to establish an independent life for themselves and their dependents (Evans, 2007; Purvin, 2007). Indeed, social security payment is important for increasing the resilience of low-income families. In the context of domestic violence, it provides material means of escape and alleviates poverty for many women ending a violent relationship (Cortis & Bullen, 2015, p. 17). However, Cortis and Bullen (2016, p. 32) report that women who have been affected by domestic violence are twice as likely to experience difficulty accessing welfare services compared to women who had not experienced violence. In Australia, access to social security payment is determined by a couple’s joint income and assets. In the context of domestic violence, a victim/survivor, if married or considered a member of a couple, will have her income and assets assessed jointly with those of the perpetrator’s. The couple rule in Australian social security law thus ties women’s access to social security payment to the income and assets of the perpetrator (Sleep, 2016; 2017). Further, when applying the couple rule, reports of domestic violence made to police or other services can be interpreted as evidence of a relationship (Sleep, Tranter & Stannard, 2006). This means that information that has been collected to enhance women’s safety is being used to assess their entitlement for social security payment. When a couple rule assessment is undertaken in the context of a victim/survivor of domestic violence attempting to redefine their relationship as part of the process of separating from the perpetrator, this can have implications for both the victim/survivor and their dependent’s physical safety and financial security. This research project adopts the definition of domestic violence used in The National Plan to Reduce Violence against Women and their Children 2010-2022 (Council of Australian Governments, 2011, p. 2), that “domestic violence includes physical, sexual, emotional and physiological abuse”. Domestic violence has been interpreted broadly to include violence experienced in domestic relationships. Almost all violence in this study was perpetrated by a current or past intimate partner; however, other forms of family violence, including child sexual abuse and elder abuse were included in this study. The term domestic violence has been used throughout this study to represent the predominance of intimate partner violence in the AAT sample, and its particular challenges in the context of the couple rule. However, the term is to be interpreted broadly to include other forms of family violence because this was also present in a number of decisions and was part of the power and control context of the decisions. This research explores the dynamic between domestic violence, social security payment and the couple rule. It examines the pre-existing data set of AAT decisions of couple rule matters and will compare this to similar decisions made in New Zealand where domestic violence can be interpreted as evidence against an ongoing relationship. These data sources provide rich, reliable and non-invasive qualitative detail of social security decision-making. This research is located in the social security law, domestic violence and social policy literature. A review of the relevant literature will be presented in four parts. First, the centrality of the couple rule in Australian social security payment provision will be outlined, as well as the rule’s heritage as a development of the historical cohabitation rule. In this section, the need for more research on the current expression of the rule will also be explored. Second, the inextricable interconnections between the couple rule and domestic violence will be identified. Third, the need for more research on the interconnection between the couple rule and domestic violence, particularly from intersectional, international and intra-national perspectives, is demonstrated. Finally, the research questions and methods are detailed.