It is now that there are two isozymes of prostaglandin endoperoxide (PGH) synthase (cyclooxygenase) called PGH synthase-1 and -2 or COX I and II. Both isozymes catalyze the conversion of arachidonate to PGH2, the committed step in the formation of both prostacyclin and thromboxane A2. PGH synthase-1 is present in platelets and endothelial cells whereas PGH synthase-2 has been detected in endothelial cells treated with cytokines and phorbol esters. PGH synthase-1 (PGHS-1) has long been thought to be the site of action of nonsteroidal anti-inflammatory drugs (NSAIDs). However, it is now clear that the second isozyme, PGH synthase-2 (PGHS-2), is also inhibited by these compounds. Cloning of the cDNAs for murine PGHS-1 and PGHS-2 has allowed us to express these two enzymes in cos-1 cells and to compare the relative sensitivities of these enzymes in vitro using a series of common NSAIDs. NSAIDs such as indomethacin, piroxicam, and sulindac sulfide preferentially inhibit PGHS-1. Ibuprofen and meclofenamate inhibit both enzymes with comparable potencies. 6-Methoxy-2-naphthylacetic acid, the active metabolite of Relafen, inhibits murine PGHS-2 preferentially. Aspirin irreversibly inhibits PGHS-1, preventing this isozyme from forming any product; in contrast, aspirin treatment of PGHS-2 causes this enzyme to form 15-hydroxy-5c,8c,11c,13t-eicosatetraenoic acid (15-HETE) instead of PGH2. These results establish that the two mouse enzymes are pharmacologically distinct and suggest that it will be possible to identify or design compounds that are completely selective for each PGHS isozyme. Because PGHS-2 is usually only expressed in inflamed tissue or after exposure to mediators of inflammation, a selective inhibitor of this isoenzyme may exhibit anti-inflammatory activity without effects on PGHS-1 of platelets and perhaps without ulcerogenic effects associated with commonly available NSAIDs.