This article explores the challenges of providing a quantitative measure of domestic violence and abuse (DVA), illustrated by the Composite Abuse Scale, a validated multidimensional measure of frequency and severity of abuse, used worldwide for prevalence studies and intervention trials. Cognitive "think-aloud" and qualitative interviewing with a sample of women who had experienced DVA revealed a tendency toward underreporting their experience of abuse, particularly of coercive control, threatening behavior, restrictions to freedom, and sexual abuse. Underreporting was linked to inconsistency and uncertainty in item interpretation and response, fear of answering truthfully, and unwillingness to identify with certain forms of abuse. Suggestions are made for rewording or reconceptualizing items and the inclusion of a distress scale to measure the individual impact of abuse. The importance of including qualitative methods in questionnaire design and in the interpretation of quantitative findings is highlighted.