Rats were trained on a nonmatching-to-sample task with delays of 2, 5, and 10 min. Subsequently, performance was assessed in three groups of rats following treatment with saline or diazepam (2.0 mg/kg) administered acutely or tested chronically in six administrations. Relative to treatment with saline, diazepam produced a deficit in discrimination performance, which was greater in the acutely treated rats than in those treated chronically. The deficit was not dependent on the length of the delays. Diazepam-treated animals differed from controls in erring on trials in which they failed to investigate both test objects, failed to investigate the test object for a long enough period of time, and displaced the test object on the preferred side of the apparatus. The hypothesis that these effects represented a sedation-like reduction in behavioral variability was also supported by evidence of a diazepam-induced decrease in gross bodily activity, increase in inactivity, and increase in latencies to respond to objects. No support was found for the involvement of diazepam-induced changes in habituation, extinction, or reward effects.