Our interest in this area has been to develop novel selective inhibitors of cyclooxygenase that have improved therapeutic properties relative to currently used NSAIDs. Toward that goal, we investigated a series of 1,2-diaryl-substituted cyclopentenes based on the hypothesis that the heterocyclic ring in DuP 697 and SC-58125 primarily provides necessary double bond geometry and could be replaced by a carbocycle (cyclopentene). 1-[2-(4-Fluorophenyl)cyclopenten-1-yl]-4-[methylsulfonyl]benzene (la) was synthesized from commercially available 1,2-dibromocyclopentene via three steps (Scheme 1). La was a very potent COX-2 inhibitor (IC50 = 0.026 μM) with essentially no COX-1 activity (IC50 > 100 μM, selectivity > 3800). A structure-activity relationship (SAR) study varying substituents on the 4-fluorophenyl ring of la showed replacing fluorine with chlorine (Ic) or methyl (Id) increased COX-2 potency (IC50 = 0.003 μM) and selectivity (> 33,000), while removing the substituent (le) reduced potency by almost two orders of magnitude. A second SAR study on geminal substitution at the 4-position of the cyclopentene ring (Table 2) revealed analogs generally had reduced selectivity due to increased COX-1 activity and were highly sensitive to steric bulk (e.g., replacing methyl with ethyl in 7a vs 7d led to over 3 orders of magnitude loss in COX-2 activity). In vivo testing showed la was orally active in the rat adjuvant-induced arthritis model (ED50 = 1.7 mpk), and no gastric lesions were observed in mice at 600 mpk or intestinal damage in rats at 200 mpk. In summary, novel 1,2-diarylcyclopentenes are very potent COX-2 inhibitors with IC50 in the low nanomolar range and enzyme selectivity ratios as high as 4 orders of magnitude, orally active in arthritis models with minimal gastrointestinal toxicity.