Aspirin, nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), and specific cyclooxygenase-2 (COX-2) inhibitors each have distinctive effects on COX-1-mediated thromboxane biosynthesis, the major determinant of platelet aggregation. It is unclear whether these effects are associated with differences in thrombogenic risks. To compare the risk for thrombotic cardiovascular events among patients receiving rofecoxib, nonselective NSAIDs, and placebo, cardiovascular safety was assessed in 5,435 participants in 8 phase IIB/III osteoarthritis trials. The median treatment exposure was 31/2 months. The primary end point assessed was the risk of any arterial or venous thrombotic cardiovascular adverse event (AE). A second analysis assessed differences in the Anti-Platelet Trialists' Collaboration (APTC) events, a cluster end point that consists of the combined incidence of (1) cardiovascular, hemorrhagic, and unknown death; (2) myocardial infarction; and (3) cerebrovascular accident. Similar rates of thrombotic cardiovascular AEs were reported with rofecoxib, placebo, and comparator nonselective NSAIDs (ibuprofen, diclofenac, or nabumetone). In trials that compared rofecoxib with NSAIDs, the incidence of thrombotic cardiovascular AEs was 1.93/100 patient-years in the rofecoxib treatment group compared with 2.27/100 patient-years in the combined nonselective NSAID group. In trials that compared rofecoxib with placebo, the incidence of thrombotic cardiovascular AEs was 2.71/100 patient-years in the rofecoxib group compared with 2.57/100 patient-years in the placebo group. Consistent with the risks of cardiovascular AEs, similar rates of APTC events were reported with rofecoxib, placebo, and comparator nonselective NSAIDs. Thus, in the rofecoxib osteoarthritis development program, there was no difference between rofecoxib, comparator nonselective NSAIDs, and placebo in the risks of cardiovascular thrombotic events.