The present analyses examined genetic influences on alcohol metabolism and their possible relationship to risk of alcohol dependence. Subjects were 206 Australian twin pairs who participated in an alcohol challenge protocol in 1979-1981, in which they were given a 0.75 g/kg dose of alcohol; blood alcohol concentrations (BACs) measured at five times over a 3-hr period after alcohol ingestion were examined. Structural equation modeling, fitting a combined autoregressive and common factor model, indicated significant heritabilities for both men and women (h2 range = 0.19-0.71), with significant parameter heterogeneity as a function of gender. In 1992-1993, both twins from 159 of the alcohol challenge pairs completed a telephone-administered psychiatric diagnostic interview. Repeated-measures MANOVAs were used to examine whether respondent's or cotwin's DSM-III-R alcohol dependence status, or parental history of alcohol problems, was associated with variation in alcohol metabolism. There was some evidence that individuals at increased genetic risk of alcohol dependence [with either a paternal history of alcohol problems (women) or an MZ male cotwin who reported a history of alcohol dependence by 1992-1993] showed lower initial BACs than other groups. However, this effect was not seen in those who themselves had a history of alcohol dependence by interview follow-up, perhaps because this relationship was already masked by a history of excessive drinking at baseline.