Background: Rifampicin (RIF) induces UGT1A1, an enzyme involved in raltegravir (RAL) elimination, thereby potentially lowering RAL exposure. We examined the pharmacokinetics of RAL in human immunodeficiency virus (HIV)-infected patients on RIF-based antitubercular therapy in the French National Agency for HIV/AIDS and Viral Hepatitis Research 12 180 Reflate Tuberculosis trial. Methods: Patients started RAL in combination with tenofovir disoproxil fumarate and lamivudine after initiation of RIF (10 mg/kg/day). In arm 1 (n = 21), they received 400 mg RAL twice daily; in arm 2 (n = 16), they received RAL 800 mg twice daily initially then 400 mg twice daily 4 weeks after RIF discontinuation. Pharmacokinetic sampling was performed over 12-hour periods, 4 weeks after initiation of RAL together with RIF (period 1), 4 weeks after RIF discontinuation (period 2), and after the RAL dose reduction in arm 2 (period 3). Results: In arm 1, the geometric mean ratio (GMR) between period 1 and period 2 was 0.94 (90% confidence interval [CI], .64-1.37) for the 12-hour area under the time-concentration curve (AUC0-12), and 0.69 (90% CI, .42-1.13) for the concentration at 12 hours (C12). In arm 2, the corresponding GMRs were 0.75 (90% CI, .48-1.17) and 1.10 (90% CI, .61-2.00) for period 1 vs period 2, and 1.10 (90% CI, .78-1.55) and 1.68 (90% CI, .88-3.23) for period 1 vs period 3. Conclusions: The double dose of RAL overcompensated for RIF induction, but the standard dose was associated with only small decreases in AUC0-12 and C12 during RIF coadministration, warranting further evaluation in patients with HIV/tuberculosis coinfection.