The authors assessed the biphasic effects of alcohol on human physical aggression. Sixty male social drinkers were assigned to 1 of 4 groups: alcohol ascending limb (AAL), alcohol descending limb (ADL), or 1 of 2 sober control groups. Aggression was assessed in the AAL and ADL groups at respective ascending or descending blood alcohol concentrations (BAC) of 0.08%. Each participant in the control groups was respectively yoked with a participant in either the AAL or the ADL group to control for the longer period of time needed to reach a BAC of 0.08% on the descending limb compared with the ascending limb (i.e., passage of time effect). The authors measured aggression using a modified version of the Taylor aggression paradigm (S. Taylor, 1967), in which electric shocks are received from and administered to a fictitious opponent during a competitive task. The AAL group was more aggressive than the ADL groups. There were no differences between the ADL group and the control groups, which suggests that alcohol does not appear to increase aggression on the descending limb. The control groups did not differ in aggression, thus ruling out a passage of time effect.