Interaction between ketoconazole, itraconazole, and midazolam was investigated in a double-blind, randomized crossover study of three phases at intervals of 4 weeks. Nine volunteers were given either 400 mg ketoconazole, 200 mg itraconazole, or matched placebo orally once daily for 4 days. On day 4, the subjects ingested 7.5 mg midazolam. Plasma samples were collected and psychomotor performance was measured. Both ketoconazole and itraconazole increased the area under the midazolam concentration-time curve from 10 to 15 times (p < 0.001) and mean peak concentrations three to four times (p < 0.001) compared with the placebo phase. In psychomotor tests (e.g., the Digit Symbol Substitution Test), the interaction was statistically significant (p < 0.05) until at least 6 hours after drug administration. Inhibition of the cytochrome P450IIIA by ketoconazole and itraconazole may explain the observed pharmacokinetic interaction. Prescription of midazolam for patients receiving ketoconazole and itraconazole should be avoided.