The present research examined the effect of alcohol aggression expectancies and subjective evaluations of alcohol's effects on aggression in intimate partner violence (IPV) perpetration among college students. We were interested in determining the extent to which these relationships differed across gender.A total of 780 (57.3% female) incoming heavy drinking college freshmen who were between the ages of 18 and 25 years completed self-reported measures of IPV perpetration, alcohol use and problems, and alcohol aggression expectancies and subjective evaluations of those expectancies as part of the baseline assessment for a larger social norms alcohol intervention study. Analyses evaluated the effect of alcohol aggression expectancies and subjective evaluations of those expectancies on IPV perpetration.Results indicated that problem drinking was positively associated with IPV perpetration for those who were lower (beta = .32, p < .001) versus those who were higher (beta = .07, p = ns) in alcohol aggression expectancies. Among men, there was a significantly stronger relationship between problem drinking and IPV perpetration among those who evaluated alcohol's effects on aggression more favorably (beta = .41, p < .001) versus less favorably (beta = .11, p = ns). Among women, there was not a significantly stronger relationship between problem drinking and IPV perpetration at less favorable (beta = .17, p < .05) versus more favorable (beta = .11, p < .06) evaluations of alcohol's effects on aggression.Findings suggest that, in understanding IPV perpetration, it may not be sufficient to evaluate expected alcohol effects without also including whether those effects are viewed as good or bad. Findings also suggest that the relationship between alcohol problems and IPV perpetration may be stronger and more straightforward for men than for women.