We compared the variability in degree and recovery from steady-state inhibition of cyclooxygenase (COX)-1 and COX-2 ex vivo and in vivo and platelet aggregation by naproxen sodium at 220 versus 440 mg b.i.d. and low-dose aspirin in healthy subjects. Six healthy subjects received consecutively naproxen sodium (220 and 440 mg b.i.d.) and aspirin (100 mg daily) for 6 days, separated by washout periods of 2 weeks. COX-1 and COX-2 inhibition was determined using ex vivo and in vivo indices of enzymatic activity: 1) the measurement of serum thromboxane (TX)B(2) levels and whole-blood lipopolysaccharide-stimulated prostaglandin (PG)E(2) levels, markers of COX-1 in platelets and COX-2 in monocytes, respectively; 2) the measurement of urinary 11-dehydro-TXB(2) and 2,3-dinor-6-keto-PGF(1alpha) levels, markers of systemic TXA(2) biosynthesis (mostly COX-1-derived) and prostacyclin biosynthesis (mostly COX-2-derived), respectively. Arachidonic acid (AA)-induced platelet aggregation was also studied. The maximal inhibition of platelet COX-1 (95.9 +/- 5.1 and 99.2 +/- 0.4%) and AA-induced platelet aggregation (92 +/- 3.5 and 93.7 +/- 1.5%) obtained at 2 h after dosing with naproxen sodium at 220 and 440 mg b.i.d., respectively, was indistinguishable from aspirin, but at 12 and 24 h after dosing, we detected marked variability, which was higher with naproxen sodium at 220 mg than at 440 mg b.i.d. Assessment of the ratio of inhibition of urinary 11-dehydro-TXB(2) versus 2,3-dinor-6-keto-PGF(1alpha) showed that the treatments caused a more profound inhibition of TXA(2) than prostacyclin biosynthesis in vivo throughout dosing interval. In conclusion, neither of the two naproxen doses mimed the persistent and complete inhibition of platelet COX-1 activity obtained by aspirin, but marked heterogeneity was mitigated by the higher dose of the drug.