Several reports have suggested an association between cytomegalovirus infection and the subsequent development of cardiac allograft vasculopathy. The difficulties in interpreting these studies include the variety of methods used for the diagnosis of cytomegalovirus infection and variable criteria for the diagnosis of cardiac allograft vasculopathy. To determine whether specific aspects of cytomegalovirus infection are risk factors for cardiac allograft vasculopathy, the patient population of the Oregon Cardiac Transplant Program was analyzed for the following variables: cytomegalovirus infection, primary cytomegalovirus infection, and persistent cytomegalovirus infection for 4 or 6 months documented by either blood or urine cultures and persistent cytomegalovirus viremia for 4 months. In the 129 patients available for analysis, there was no higher incidence of cardiac allograft vasculopathy in patients with or without cytomegalovirus infection, nor was there a higher incidence of cardiac allograft vasculopathy in primary cytomegalovirus infection. There was a nonstatistically significant trend toward an increased incidence of cardiac allograft vasculopathy in patients with persistent cytomegalovirus infection as assessed by cultures positive for infection in either blood or urine. There was, however, a significant increase in the incidence of cardiac allograft vasculopathy in patients who had persistent viremia for at least 4 months compared with those without this finding (47% vs 18%, respectively; p = 0.012). In our population persistent cytomegalovirus viremia and presumably long-term exposure of the allograft coronary tree to cytomegalovirus is associated with cardiac allograft vasculopathy.