The marked analgesic efficacy of ketorolac in humans, relative to other nonsteroidal anti-inflammatory drugs (NSAIDs), has lead to speculation as to whether additional non-NSAID mechanism(s) contribute to its analgesic actions. To evaluate this possibility, we characterized (R,S)-ketorolac's pharmacological properties in vivo and in vitro using the nonselective cyclooxygenase (COX) inhibitors [indomethacin (INDO) and diclofenac sodium (DS)] as well as the selective COX-2 inhibitor, celecoxib, as references. The potency of racemic (R,S)-ketorolac was similar in tests of acetic acid-induced writhing, carrageenan-induced paw hyperalgesia, and carrageenan-induced edema formation in rats; ID50 values = 0.24, 0. 29, and 0.08 mg/kg, respectively. (R,S)-ketorolac's actions were stereospecific, with (S)-ketorolac possessing the biological activity of the racemate in the above tests. The analgesic potencies for (R,S)-, (S)-, and (R)-ketorolac, INDO, and DS were highly correlated with their anti-inflammatory potencies, suggesting a common mechanism. (R,S)-ketorolac was significantly more potent than INDO or DS in vivo. Neither difference in relative potency of COX inhibition for (R,S)-ketorolac over INDO and DS nor activity of (S)-ketorolac at a number of other enzymes, channels, or receptors could account for the differences in observed potency. The distribution coefficient for (R,S)-ketorolac was approximately 30-fold less than for DS or INDO, indicating that (R,S)-ketorolac is much less lipophilic than these NSAIDs. Therefore, the physicochemical and pharmacokinetics properties of (R,S)-ketorolac may optimize the concentrations of (S)-ketorolac at its biological target(s), resulting in greater efficacy and potency in vivo.