The effect of rifampicin and isoniazid pretreatment on the pharmacokinetics of quinine after a single oral dose (600 mg quinine sulphate) was studied in nine healthy young Thai male volunteers using a three-way randomized crossover design. Subjects were studied over three 2 day periods, during which they received no pretreatment, or pretreatment with daily 600 mg p.o. rifampicin for 2 weeks, or isoniazid 300 mg p.o. daily for 1 week, prior to quinine administration. The mean (+ s.d.) clearance (CL/F) of quinine coadministered with rifampicin (0.87 ± 0.35 1 h-1 kg-') was significantly greater than that of quinine alone (0.14 ± 0.05 1 h-1 kg-'). The mean difference in clearance from the control treatment was 0.73 1 h-1 kg-', with 95% confidence interval (C.I.) of 0.48 to 0.98. The unbound clearance (CLu/F) of quinine, which reflects the activity of the drug-metabolizing enzymes, was considerably greater (6.9-fold) in subjects when rifampicin was coadministered with quinine than that of quinine alone (6.9 ± 3.6 vs 1.0 ± 0.5 1 h-1 kg-'; the 95% C.I. for the mean difference was 3.3 to 8.5). The mean elimination half-life of quinine when coadministered with rifampicin (5.5 ± 3.0 h) was significantly shorter than when quinine was given alone (11.1 ± 3.0 h; the 95% C.I. for the mean difference was -8.6 to -2.6). In contrast to rifampicin, pretreatment for 1 week with 300 mg oral isoniazid had no significant effects on the pharmacokinetics of quinine. These results indicate that rifampicin pretreatment caused a marked increase (6.2-fold) in the clearance of quinine, possibly due to enzyme induction. The extent to which the elimination of quinine is enhanced by the concomitant administration of rifampicin is likely to have important clinical consequences. Although the clinical significance of these findings is unknown, they indicate the need for caution in the administration of quinine to patients who are concurrently taking rifampicin as an anti-tuberculosis medication.