Measurement of domestic violence in population surveys has generated significant debate over a number of years, with researchers and practitioners split over the extent to which violence in intimate relationships is ‘gendered’. While many surveys suggest women and men experience similar levels of domestic violence, domestic violence practitioners see abuse as perpetrated primarily by men against women. Those writing from a feminist-advocate perspective propose that ‘gender symmetrical’ headline survey prevalence estimates are misleading and reflect problems with existing measurement frameworks. Two potential solutions have emerged to this problem: measuring acts of physical violence and associated criminal offences more accurately, and measuring the non-physically abusive context in which many acts of violence are situated. This article proposes that these two solutions are not incompatible, and that, in fact, both are desirable in order to illuminate fully the gendered nature of intimate violence. But while counting the frequency and severity of physical violence committed by men and women will go some way towards foregrounding gender in this debate, measuring non-violent coercion that falls outside traditional crime codes is fundamental, both to highlighting the true nature and severity of abuse experienced by women and men and identifying primary victims and perpetrators.