1. We have evaluated the selectivity of ketoprofen and two novel nonsteroidal anti-inflammatory drugs, N-[2-(cyclohexyloxy)-4-nitrophenyl]methanesulphonamide (NS-398) and 5-methanesulphonamido-6-(2,4-difluorothiophenyl)-1-indano ne (L-745,337), in inhibiting the cyclo-oxygenase activity of prostaglandin endoperoxide synthase-2 (PGHS-2) vs PGHS-1 in human blood monocytes and platelets, respectively. 2. Heparinized whole blood samples were drawn from healthy volunteers pretreated with aspirin, 300 mg 48 h before sampling, to suppress the activity of platelet PGHS-1 and incubated at 37 degrees C for 24 h with increasing concentrations of the test compounds in the presence of lipopolysaccharide (LPS, 10 micrograms ml-1). Immunoreactive PGE2 levels were measured in plasma by a specific radioimmunoassay as an index of the cyclo-oxygenase activity of LPS-induced monocyte PGHS-2. 3. The effects of the same inhibitors on platelet PGHS-1 activity were assessed by allowing whole blood samples, drawn from the same subjects in aspirin-free periods, to clot at 37 degrees C for 1 h in the presence of the compounds and measuring immunoreactive thromboxane B2 (TXB2) levels in serum by a specific radioimmunoassay. 4. Under these experimental conditions, ketoprofen enantioselectively inhibited the cyclo-oxygenase activity of both PGHS-1 and PGHS-2 with equal potency (IC50 ratio: approx. 0.5 for both enantiomers), while L-745,337 and NS-398 achieved selective inhibition of monocyte PGHS-2 (IC50 ratio: > 150). L-745,337 and NS-398 did not affect LPS-induced monocyte PGHS-2 biosynthesis to any detectable extent. 5. We conclude that L-745,337 and NS-398 are selective inhibitors of the cyclo-oxygenase activity of human monocyte PGHS-2. These compounds may provide adequate tools to test the contribution of this novel pathway of arachidonate metabolism to human inflammatory disease.