It has been proposed that cyclooxygenase (COX)-1 and COX-2 subserve different physiologic functions largely because of the striking differences in their tissue expression and regulation. COX-1 displays the characteristics of a "housekeeping" gene and is constitutively expressed in almost all tissues. COX-1 appears to be responsible for the production of prostaglandins (PG) that are important for homeostatic functions, such as maintaining the integrity of the gastric mucosa, mediating normal platelet function, and regulating renal blood flow. In sharp contrast, COX-2 is the product of an "immediate-early" gene that is rapidly inducible and tightly regulated. Under basal conditions, COX-2 expression is highly restricted; however, COX-2 is dramatically upregulated during inflammation. For example, synovial tissues in patients with rheumatoid arthritis (RA) express increased levels of COX-2. In animal models of inflammatory arthritis, COX-2 increases in parallel with PG production and clinical inflammation. In vitro experiments have revealed increased COX-2 expression after stimulation with proinflammatory cytokines, such as interleukin 1 (IL-1) and tumor necrosis factor-alpha (TNF-alpha), in many cell types, including synoviocytes, endothelial cells, chondrocytes, osteoblasts, and monocytes/macrophages. Another distinguishing characteristic of COX-2 is decreased expression in response to glucocorticoids. COX-2 is also increased in some types of human cancers, particularly colon cancer. Mechanisms underlying the association between COX-2 overexpression and tumorigenic potential may include resistance to apoptosis, or programmed cell death. Upregulated COX-2 expression undoubtedly plays a role in pathologic processes characterized by increased local PG production. One would predict, based on current information regarding the differential tissue expression of COX-1 and COX-2, that highly selective inhibitors of COX-2 will provide effective antiinflammatory activity with marked reduction in toxicity.