A single dose of racemic ketorolac (30 mg of the tromethamine salt, Toradol®) was administered by bolus intramuscular injection to four young, healthy volunteers. The concentrations of total (bound plus unbound) (R)-and (S)-ketorolac were measured in plasma for 9 h after dosing. The mean ± s.d. clearance of (S)-ketorolac (45.9 ± 10.1 ml h-1 kg-') exceeded (P = 0.0032) that of the (R)-enantiomer (19.0 ± 5.0 ml h-1 kg-1). The mean ± s.d. AUC ratio for (S)-ketorolac:(R)-ketorolac (0.442 ± 0.043) was signifcantly different from unity (P = 0.0001). The steady-state volume of distribution of (S)-ketorolac (0.135 ± 0.0221 kg-1) was significantly different (P = 0.0013) from that of its optical antipode (0.075 ± 0.014 1 kg-') and the half-lives of (S)-and (R)-ketorolac (2.35 ± 0.23 h and 3.62 ± 0.79 h, respectively) were also significantly different (P = 0.026). These data indicate that the disposition of ketorolac in man is subject to marked enantioselectivity and, because of possible differences in biological activity of (S)-and (R)-ketorolac, emphasize the need to monitor separate stereoisomer concentrations of the drug if pharmacological data are to be interpreted correctly.