One hundred two healthy men were evaluated in one of three studies conducted to evaluate the coadministration of nefazodone, 200 mg twice daily, and three benzodiazepines: triazolam, 0.25 mg; alprazolam, 1 mg twice daily; or lorazepam, 2 mg twice daily. In the first study, psychomotor performance, memory, and sedation were assessed at 0, 0.5, 1.5, 2.5, and 9 hours after single doses of triazolam alone and again after 7 days of nefazodone. Data from 6 of 12 subjects in this study were evaluable because of a dosing error in the other 6 subjects. In the subsequent two parallel design studies, groups of 12 volunteers received 7 days of either placebo; nefazodone, 200 mg; alprazolam, 1 mg twice daily; or alprazolam plus nefazodone or, in the second study, either placebo; nefazodone; lorazepam, 2 mg twice daily; or lorazepam plus nefazodone; the studies were identical, double-dummy, double-blind designs. Psychomotor performance, memory, and sedation were assessed at 0, 1, 3, and 8 hours after the 8 a.m. dose on days 1, 3, 5, and 7 of the studies. In all studies, blood samples were also obtained at testing times so that effect/concentration comparisons could be made and so full pharmacokinetic analyses could be done for separate studies. Nefazodone had no effect on psychomotor performance, memory, or sedation relative to placebo in any study. The mean maximum observed effect (MaxOE) on psychomotor performance and sedation were increased when triazolam was given after 7 days of nefazodone (p < 0.05); also, triazolam concentration was 60% higher at this time. Alprazolam and lorazepam impaired performance on day 1 (mean MaxOE, 34 and 30%, respectively) relative to placebo and nefazodone. By day 7 of alprazolam or lorazepam, psychomotor impairment decreased, indicating the development of tolerance. Alprazolam plus nefazodone increased psychomotor impairment (MaxOE, approximately 50%) and sedation relative to alprazolam alone on days 3, 5, and 7 (p < 0.05). Higher alprazolam concentrations explained the increased impairment in the alprazolam plus nefazodone treatment group; however, it is also possible that there was a delay in the development of tolerance. There were no differences in psychomotor impairment, memory, sedation, or lorazepam concentration detected between the lorazepam alone and lorazepam plus nefazodone treatments. This is consistent with the absence of a pharmacokinetic interaction between nefazodone and lorazepam. These results indicate that if the coadministration of a benzodiazepine is required in patients receiving nefazodone therapy, clinically significant interactions would be less likely with those eliminated by conjugative metabolism such as lorazepam. In cases where a benzodiazepine eliminated by oxidative metabolism is required, a reduction in initial dosage and careful clinical evaluation for signs of psychomotor impairment may be appropriate.