Some nonsteroidal anti-inflammatory drugs (NSAIDs), including cyclooxygenase-2 selective inhibitors, have been associated with increased cardiovascular (CV) events in recent clinical trials or observational studies. To determine whether the cyclooxygenase-2 selective inhibitor celecoxib affects CV risk, the incidence of CV events was analyzed in patients treated with celecoxib, placebo, or nonselective NSAIDs in the clinical trial database for celecoxib using defined Antiplatelet Trialists' Collaboration end points of nonfatal myocardial infarction, nonfatal stroke, and CV death. Patient data were derived from studies in osteoarthritis, rheumatoid arthritis, ankylosing spondylitis, low back pain, and Alzheimer's disease. This meta-analysis included (1) 7,462 patients exposed to celecoxib 200 to 800 mg/day for 1,268 patient-years compared with 4,057 patients treated with placebo for 585 patient-years, and (2) 19,773 patients treated with celecoxib 200 to 800 mg/day for 5,651 patient-years compared with 13,990 patients treated with nonselective NSAIDs (diclofenac, ibuprofen, naproxen, ketoprofen, and loxoprofen) for 4,386 patient-years. CV events were adjudicated by a 3-member expert end point committee (WBW, JSB, PBG) blinded to treatment group and study. The incidence rates of the combined CV events were not significantly different between patients treated with celecoxib and placebo or between those treated with celecoxib and nonselective NSAIDs. Event rates were similar for adjudicated and nonadjudicated data. Dose of celecoxib, the use of aspirin, or the presence of CV risk factors did not alter these results. In conclusion, these analyses failed to demonstrate an increased CV risk with celecoxib relative to placebo and demonstrated a comparable rate of CV events with celecoxib treatment compared with nonselective NSAIDs.