Before the withdrawal of 2 COX-2 selective agents (COX-2s) from the market, many rheumatoid arthritis patients were using these products regularly, with disease-modifying antirheumatic agents. Clinical trials have shown benefit of COX-2s equivalent to nonselective nonsteroidal anti-inflammatory drugs (NS-NSAIDs) in rheumatoid arthritis. Better gastrointestinal (GI) safety has been demonstrated with COX-2s; numerical but not statistical benefit with concomitant use of cardiovascular (CV) doses of aspirin. COX-2 benefit may extend to lower GI blood loss against which proton pump inhibitors are not protective. COX-2s are associated with hypertension and edema of similar magnitude to NS-NSAIDs in predisposed individuals. Epidemiologic studies and clinical trials have confirmed the association of serious thromboembolic (CV) events and congestive heart failure with rofecoxib >25 mg daily, celecoxib, and NS-NSAIDs, although there is a paucity of long-term data. Important questions remain regarding relative GI and CV risks: is concomitant aspirin protective when coadministered with COX-2s? Does this abrogate their GI benefit? As identified many years ago with NS-NSAIDs, patients may respond to one and not another; COX-2s should be considered individually and not as a single class. Patients deserve the opportunity to make a choice about the perceived benefit/risk assessment when using these therapies, with the collaboration of their physician.