Although alcohol-related violence is a common problem, there is wide variability amongst individuals in the extent to which they engage in alcohol-related violence. To assist with identification of individual differences, we developed a comprehensive measure of alcohol-related aggression. Items for the Alcohol-related Aggression Questionnaire (ARAQ) were generated from a priori literature searches and administered to 226 males. Factor analysis reduced the ARAQ from 64 to 28 items, with five subscales showing good internal consistency and concurrent validity. In a second study, 219 offenders and non-offenders were tested on the ARAQ-28, the Buss-Perry Aggression Questionnaire and the Alcohol Use Disorders Identification Test. Confirmatory factor analysis did not support the original factor structure, but an exploratory factor analysis gave a new four-factor structure with broadly similar constructs. The ARAQ-28 showed good internal consistency, test-retest reliability and construct and discriminant validity. The ARAQ-28 coheres as a univariate scale, and its subscales may have clinical utility.