One hundred five patients underwent mitral valve replacement for relief of isolated mitral regurgitation between 1974 and 1979. There were 4 in-hospital deaths (4 percent) and 12 late deaths giving an 82 percent predicted 5 year survival rate. An age of 60 years or more at the time of surgery and a preoperative left ventricular ejection fraction of less than 0.40 were the only variables that correlated with decreased survival at 3 to 5 years after operation (p< 0.05). Postoperatively, 87 (98 percent) of 89 long-term survivors were in New York Heart Association functional class I or II (68 in class I and 19 in class II). Survival did not differ between patients with porcine versus mechanical valve replacement, but patients with a mechanical valve had a greater incidence of postoperative cerebrovascular accident (8.6 100 patient years) than did patients with a porcine valve (2.8/100 patient years)(p< 0.002). Ejection fraction at rest was determined with multigated cardiac imaging 12 to 75 months post-operatively in 34 of 89 long-term survivors. The mean preoperative ejection fraction was 0.62±0.09 (mean±1 standard deviation) and the mean postoperative ejection fraction was 0.50±0.15 (p< 0.001). When the preoperative value was compared with the postoperative value at rest the ejection fraction increased by 0.10 or more in 1 patient (3 percent), remained within±0.09 of the preoperative value in 12 patients (35 percent) and decreased by 0.10 or greater in 21 patients (62 percent). Sixteen (94 percent) of 17 patients whose postoperative ejection fraction was greater than 0.50 were in functional class I postoperatively compared with 11 (65 percent) of 17 patients whose postoperative ejection fraction was 0.50 or less (p< 0.05). No preoperative factor, including preoperative ejection fraction or cardiothoracic ratio, predicted the postoperative ejection fraction. A postoperative exercise ejection fraction was obtained in 29 patients, and an abnormal ejection fraction change with exercise (increase< 0.05) was observed in 20 patients (69 percent). Patient age at the time of study correlated inversely with the change in ejection fraction from rest to exercise; no other variables were predictive. It is concluded that, in addition to age, only preoperative left ventricular function as measured by ejection fraction predicts survival in patients undergoing mitral valve replacement for isolated mitral regurgitation. Clinical recovery is good even though the majority of long-term survivors have a postoperative decrease in ejection fraction.