Background. Interaction with platelet function by non-steroidal anti-inflammatory drugs (NSAIDs) is related to the inhibition of cyclo-oxygenase-1 (COX-1). In patients with rheumatoid arthritis (RA), only one of the COX-2-selective NSAIDs (nabumetone) has been demonstrated to spare platelet function partially. Objective. To compare the effects of the COX-2-selective inhibitor, meloxicam, with those of the non-selective NSAID, naproxen, on platelet function and thromboxane levels in RA patients. Methods. In this randomized, controlled, cross-over trial, 10 RA patients used meloxicam 7.5 mg bid and naproxen 500 mg bid, each during a 2-week period. Washout periods were applied. Before and after each 2-week period of NSAID intake, laboratory studies were performed. Results. Platelet aggregation was significantly less influenced, thromboxane levels were less inhibited (246 vs 117 pg/ml) and bleeding times were less prolonged with meloxicam than with naproxen (3.2 vs 2.3 min). Moreover, the results of all tests during meloxicam exposure were comparable with baseline values. Conclusion. In RA patients, meloxicam, a representative of the selective COX-2 inhibitors, does not interfere with platelet function and thromboxane levels, in contrast with naproxen (a non-selective COX inhibitor).