Two abuser groups (43 uncounseled and 46 counseled) drank significantly greater amounts and for different psychological reasons than two nonabuser groups (50 satisfactorily married and 42 discordant); however, group frequency rates did not vary. The abuser groups had experienced and observed significantly more violence during childhood than the nonabuser groups; also, they reported significantly more current life stress. The female partners of the abusers did not differ in global alcohol use; however, they drank substantially more than their counterparts. They were especially prone to drink in response to battering. Approximately one-fourth of the abusers and one-fifth of their wives drank frequently during abusive episodes. Stressors and depressors most likely precipitated both drinking and battering, with strain, hostility, and alcohol myopia (altered perception) serving as intervening variables. Results clarify contradictory reports concerning the role of alcohol in family violence. 6 tables and 44 references