The objective of this study was to investigate pharmacokinetic and pharmacodynamic interactions between midazolam and fluoxetine, fluvoxamine, nefazodone, and ketoconazole. Forty healthy subjects were randomized to receive one of the four study drugs for 12 days in a parallel study design: fluoxetine 60 mg per day for 5 days, followed by 20 mg per day for 7 days; fluvoxamine titrated to a daily dose of 200 mg; nefazodone titrated to a daily dose of 400 mg; or ketoconazole 200 mg per day. All 40 subjects received oral midazolam solution before and after the 12-day study drug regimen. Blood samples for determination of midazolam concentrations were drawn for 24 hours after each midazolam dose and used for the calculation of pharmacokinetic parameters. The effects of the study drugs on midazolam pharmacodynamics were assessed using the symbol digit modalities test (SDMT). The mean area under the curve (AUC) for midazolam was increased 771.9% by ketoconazole and 444.0% by nefazodone administration. However, there was no significant change in midazolam AUC as a result of fluoxetine (13.4% decrease) and a statistical trend for fluvoxamine (66.1% increase) administration. Pharmacodynamic data are consistent with pharmacokinetic data indicating that nefazodone and ketoconazole resulted in significant increases in midazolam-related cognition impairment. The significant impairment in subjects' cognitive function reflects the changes in midazolam clearance after treatment with ketoconazole and nefazodone. These results suggest that caution with the use of midazolam is warranted with potent CYP3A4 inhibitors.