This study assessed the safety, tolerability, and pharmacokinetic interaction between PA-824, a novel antitubercular nitroimidazo-oxazine, and midazolam, a CYP3A4 substrate, in 14 healthy adult male and female subjects. The study followed up on observations in vitro that PA-824 caused weak and time-dependent inhibition of CYP3A4. Subjects received a single oral dose of midazolam (2 mg), followed by a 2-day washout. After the washout, all subjects received PA-824 (400 mg) once daily for 14 consecutive days. On day 14, all subjects received the final PA-824 dose coadministered with a 2-mg oral dose of midazolam. The pharmacokinetic endpoints AUC0-t, AUC(0-∞), and C(max) for midazolam and 1-hydroxy midazolam were compared between midazolam administered alone versus midazolam coadministered with PA-824. Statistical analysis demonstrated that the mean midazolam values of C(max), AUC(0-t), and AUC(0-∞) parameters were reduced by ca. 16, 15, and 15%, respectively, when PA-824 was coadministered with midazolam. The total exposure (AUC) of 1-hydroxy midazolam was 13 to 14% greater when coadministered with PA-824 compared to midazolam administered alone. The Cmax of 1-hydroxy midazolam was similar between treatments. Based on these results, PA-824 does not inhibit or induce CYP3A4 to a clinically meaningful extent and is not likely to markedly affect the pharmacokinetics of CYP3A4 metabolized drugs.