NSAIDs have been the mainstay of treatment in the management of pain and inflammation associated with chronic inflammatory disorders. They are effective. However, complications arising from chronic NSAID use are common and are primarily due to gastrointestinal (GI) toxicity in the form of gastritis, peptic erosions and ulceration and GI bleeds. GI toxicity has been attributed to the blockade of the cyclo-oxygenase (COX)-1-mediated generation of the cytoprotective prostanoids, such as prostaglandin (PG) E2 and PGI2 (prostacyclin). More recently, selective COX-2 inhibitors ('coxibs') were designed to inhibit the production of COX-2-dependent inflammatory prostanoids and to leave intact the cytoprotective COX-1 products. The coxibs, while exhibiting similar efficacy to traditional NSAIDs in controlled clinical trials of their efficacy in chronic inflammatory conditions, such as osteoarthritis and rheumatoid arthritis, have been associated with a reduced incidence of surrogate or actual indices of GI toxicity. However, concerns regarding cardiovascular safety in high-risk patients have evolved. These concerns were driven initially by the concept that inhibition of COX-2-derived endothelial PGI2 without concomitant inhibition of platelet thromboxane A2 would result in increased cardiovascular risk. This was borne out in the Vioxx Gastrointestinal Outcomes Research study of rofecoxib, but not demonstrated in the Celecoxib Long Term Arthritis Safety Study trial. Further elucidation of the relative roles of COX-1- and COX-2-generated prostanoids has enabled a greater understanding of the biology of these pathways. However, it is still not completely clear how this understanding may be appropriately translated into clinical medicine.