Background: Selective inhibitors of cyclooxygenase-2 (COX-2) called coxibs, are effective anti-inflammatory and analgesic drugs. Recently, these drugs were associated with an increased risk for myocardial infarction and atherothrombotic events. The hypothesis of thromboxane-prostacyclin imbalance has been preferred to explain these unwanted effects., Methods: We studied the effects of 14 days intake of rofecoxib (25 mg q.d.), celecoxib (200 mg b.i.d.), naproxen (500 mg b.i.d.) and placebo in a randomized, blinded, placebo-controlled study in young healthy volunteers (median age 25-30 years, each group n = 10). We assessed prostanoid metabolite excretion (PGE-M, TXB(2), 6-keto-PGF(1alpha), 11-dehydro-TXB(2), 2,3-dinor-TXB(2), and dinor-6-keto-PGF(1alpha)), the expression of platelet activation markers (CD62P, PAC-1, fibrinogen), platelet-leukocyte formation, the endogenous thrombin potential, platelet cAMP content and plasma thrombomodulin level., Results: Naproxen suppressed biosynthesis of PGE-M, prostacyclin metabolites and thromboxane metabolites and thrombomodulin levels. In contrast, both coxibs had an inhibitory effect only on PGE-M, 6-keto-PGF(1alpha), and on dinor-6-keto-PGF(1alpha), whereas TXB(2), 2,3-dinor-TXB(2) and 11-dehydro-TXB(2) excretion were unaffected. None of the coxibs exerted significant effects on the expression of platelet activation markers, cAMP generation, platelet-leukocyte formation, or on thrombomodulin plasma levels. Interestingly, platelet TXB(2) release during aggregation was enhanced after coxib treatment following arachidonic acid or collagen stimulation., Conclusion: In young healthy volunteers coxibs inhibit systemic PGE(2) and PGI(2) synthesis. Platelet function and expression of platelet aggregation markers are not affected; however, coxibs can stimulate TXB(2) release from activated platelets. Combined decrease in vasodilatory PGE(2) and PGI(2) together with increased TXA(2) in proaggregatory conditions may contribute to coxib side effects.