Midazolam (MDZ) total clearance (ClT) is widely used for cytochrome P450 3A (CYP3A) phenotyping, but requires up to eight blood samples. This study was conducted to compare the use of midazolam ClT to use of a midazolam urinary metabolic ratio for CYP3A phenotyping. Ten male and 10 female subjects received i.v. midazolam 0.025 mg/kg eight times over a 4-month period at approximately 2-week intervals. The first six phenotyping measures were used to estimate baseline CYP3A activity, then subjects received the moderate CYP3A inhibitor fluvoxamine 150 mg/day for the last 4 weeks (two phenotyping visits) of the study. Serial blood samples were obtained for calculation of ClT. Urine was collected for 6 h following each midazolam dose. Midazolam, 1'-hydroxymidazolam (1-OHMDZ), and 4-hydroxymidazolam were measured in plasma and urine by liquid chromatography with tandem mass spectrometry (LC/MS/MS). Analysis of 148 samples from 20 subjects revealed a weak overall correlation between the urinary ratio of 1-OHMDZ/MDZ to midazolam ClT of r(s) = 0.372 (P = 0.0001). There was no correlation when examining either baseline samples or fluvoxamine-inhibited samples alone (r(s) = 0.101, P = 0.289 and r(s) = 0.266, P = 0.123, respectively). The median (range) urinary ratio decreased significantly with fluvoxamine [219 (141-409) versus 127 (50-464); P = 0.005] and to a similar extent to the midazolam ClT (-33.6% versus -42.4%, respectively; P > 0.05). Median urinary recovery of the i.v. midazolam dose varied between 1.4% and 53% and was significantly lower in samples collected while patients were receiving fluvoxamine (34.3% versus 23.1%; P= 0.0004). Based on these results, although this midazolam urinary ratio was not very reflective of baseline CYP3A activity, it may be a useful indicator of CYP3A inhibition.