Despite initial promising reports that anti-inflammatory properties of cycloxygenase-2 (COX-2) inhibitors may confer anti-atherosclerosis effects and stabilize the atherosclerotic plaque, subsequent data from long-term clinical trials have shown that selective COX-2 inhibitors are associated with increased risk of cardiovascular events. The commonly cited explanation is that selective inhibition of COX-2 leads to depletion of prostacyclin, whereas the production of pro-thrombotic thromboxane by means of cycloxygenase-1 (COX-1) is unopposed. This hypothesis seems unlikely as the overall explanation, because low-dose aspirin does not decrease the increased risk associated with COX-2 inhibitors. Moreover, the risk associated with nonselective COX inhibitors may be similar to selective COX-2 inhibitors. Alternative hypotheses include (1) elevated blood pressure, (2) abnormal vascular remodeling, (3) inhibition of protective mechanisms against ischemia-reperfusion injury, and (4) inhibition of 15-epi-lipoxin production. Varying results in different experimental models may be related to the fact that COX-2 is involved in numerous cellular functions. Inhibiting COX-2 in inflammatory cells may have favorable effects, whereas in organs such as the heart and brain and/or blood vessels may have deleterious effects. Currently, the "selective COX-2 inhibitors" are not selective in the sense that they inhibit COX-2 in all tissues without predilection to inflammatory cells and, as a result, may summate to increase the risk of cardiovascular events.