The effects of the new cyclooxygenase (COX)-2 selective inhibitor, valdecoxib (40 mg bid; n = 17), on platelet function were evaluated, along with ibuprofen (800 mg tid; n = 15) and placebo (n = 15), in healthy elderly subjects (65-85 years) in this 7.5-day, randomized, single-center, double-blind study. Platelet aggregation (to sodium arachidonate, collagen, and adenosine diphosphate), bleeding time, and serum thromboxane B2 (TxB2) concentrations were measured up to 8 hours postdose on Days 1 and 8. Valdecoxib had no platelet effects, while ibuprofen significantly decreased platelet aggregation, significantly increased bleeding time (2-4 h postdose on each day), and significantly decreased TxB2 levels at all time points. In conclusion, unlike ibuprofen, valdecoxib 40 mg bid spares platelet COX-1 function in healthy elderly subjects. Valdecoxib's lack of effect on platelet aggregation and bleeding time suggests that it will have an improved clinical profile over nonselective NSAIDs, particularly in patients for whom bleeding complications are a concern.