A positive family history of alcohol use disorders (FH) is a robust predictor of personal alcohol abuse and dependence. Exposure to problem-drinking models is one mechanism through which family history influences alcohol-related cognitions and drinking patterns. Similarly, exposure to alcohol advertisements is associated with alcohol involvement and the relationship between affective response to alcohol cues and drinking behavior has not been well established. In addition, the collective contribution that FH, exposure to different types of problem-drinking models (e.g. parents, peers) and personal alcohol use have on appraisal of alcohol-related stimuli has not been evaluated with a large sample. We investigated the independent effects of FH, exposure to problem-drinking models and personal alcohol use on valence ratings of alcohol pictures in a college sample. College students (n = 227) completed measures of personal drinking and substance use, exposure to problem-drinking models, FH and ratings on affective valence of 60 alcohol pictures. Greater exposure to non-familial problem-drinkers predicted greater drinking among college students (β = 0.17, P < 0.01). However, personal drinking was the only predictor of valence ratings of alcohol pictures (β = −0.53, P < 0.001). Personal drinking level predicted valence ratings of alcohol cues over and above FH, exposure to problem-drinking models and demographic characteristics. This suggests that positive affective responses to alcohol pictures are more a function of personal experience (i.e. repeated heavy alcohol use) than vicarious learning.