We examined the hypothesis that genetically determined differences in sensitivity to alcohol explain some of the genetic variation in alcohol consumption pattern. Self-report data on average weekly alcohol consumption and self-ratings of intoxication after a standard dose of ethanol (0.75 g/kg body weight), used as an index of sensitivity, were obtained on 206 Australian twin pairs. Significant genetic covariance between weekly consumption and level of intoxication after alcohol intake was found in males, lower ratings of intoxication being associated with increased consumption. However, when direction of causation models were fitted to the male twin data, the hypothesis that decreased sensitivity was a cause of increased consumption was rejected. The major causal effect was that of weekly consumption on level of sensitivity. A similar, although nonsignificant, trend was observed in females. The strength of the association between self-report of average weekly consumption and level of intoxication after a standard dose of alcohol supports the validity of the former measure.