One hundred twenty-two consecutive patients (104 men; 18 women) were studied to determine the incidence and natural history of pericardial effusion occurring 2, 5, 10, and 20 to 50 days after cardiac surgery. Three patients had pericardial effusions before and 103 patients (91 men; three women) had effusions after surgery. Effusions were first recorded on the second postoperative day in 72 patients, on the fifth postoperative day in 29 patients, and on the tenth postoperative day in two patients. In 96 of these patients, effusions reached their maximum size by postoperative day 10. Of the 103 patients with effusions, 66 (64%) were followed to complete resolution. A specific pattern was observed in most resolving effusions. The echo-free space diagnostic of pericardial effusion became progressively more echo-dense as the effusion diminished in size. As the effusion became echo-dense, the posterior pericardium, which had been motionless, resumed its normal systolic anterior motion. One patient developed cardiac tamponade on postoperative day 3. We conclude that pericardial effusion occurs frequently after cardiac surgery, but that associated complications are rare. Circulation 69, No. 3, 506-511, 1984. ECHOCARDIOGRAPHY is the technique of choice for the diagnosis of pericardial effusion. In 1974. Horowitz et al.' correlated the size of pericardial effusions with the degree of separation between epicardium and pericardium. They were able to accurately detect as little as 15 to 20 ml of pericardial fluid. Martin et al.2 also showed that accurate quantification of pericardial effusion could be derived from M mode echocardiograms.2 The correlation for moderate-sized effusions was best, while that for small or large effusions was less accurate. Feigenbaum3 emphasized that lack of side effects and reproducibility of results make echocardiography ideally suited for serial studies. Also, with the development of portable equipment, the bedside evaluation of critically ill patients became possible.Pericardial effusion is very common after cardiac surgery. Despite the high incidence of effusion, the clinical approach to this problem remains controversial. Prospective studies defining the course and outcome of postoperative pericardial effusion have not