The interaction of alcohol and visuospatial learning was evaluated during two experimental sessions. During one session, participants were experimentally intoxicated to obtain a blood alcohol concentration of at least 0.10%. During another session, a nonalcohol placebo was administered. The learning task consisted of a paired-associate paradigm requiring participants to learn the distinct spatial positions of 6 visually presented "nonsense shapes." The visuospatial learning of participants in the placebo condition was generally superior to their learning while intoxicated. However, intercorrelations of performance measures indicated that the relation between alcohol ingestion and performance differences across conditions was not linear.