Patients suffering from rheumatoid arthritis received oral doses of 600 mg racemic ibuprofen (n = 25; RAC) or 400 mg (S)-ibuprofen (n = 25; S-IBU) in a double-blind, randomized parallel-group study. 2. The pharmacokinetic parameters of (S)-ibuprofen were not statistically different between treatments (P > 0.05). Comparing (S)- and (R)-ibuprofen within the group receiving the racemate significantly higher Cmax (20.3 +/- 5.3 vs 17.7 +/- 4.4 micrograms ml-1; P < 0.02; 95% confidence interval for differences (CI): 0.5-4.6), AUC (86.2 +/- 23.5 vs 67.6 +/- 26.6 micrograms ml-1 h; P < 0.001; CI: 9.5-27.6), mean residence time (4.5 +/- 1.1 vs 4.1 +/- 1.2 h; P < 0.01; CI: 0.1-0.6) and renal clearance (0.8 +/- 0.6 vs 0.0 +/- 0.0 ml min-1; P < 0.001; CI: 0.5-1.0) values were observed for the (S)-enantiomer. 3. No difference was found (P > 0.05) between treatments in the percentage of the dose recovered in the urine as (R)- or (S)-ibuprofen plus metabolites (S-IBU: 80.2 +/- 8.47 vs RAC: 74.1 +/- 14.0%). 4. Interindividual variation in the pharmacokinetics of (S)-ibuprofen following administration of the racemate was similar to that following the administration of the single isomer suggesting that chiral inversion is not a major factor contributing to variability in the disposition of this drug.