Functional capacity is an important outcome variable for heart transplantation, but there are few data that examine the temporal relation and duration of improvement in cardiopulmonary exercise performance after cardiac transplantation. Cardiopulmonary exercise performance was measured in 140 patients who underwent 426 treadmill exercise tests up to 9 years after cardiac transplantation. Univariate and multivariate analyses were used to predict postoperative improvement in functional capacity. Peak oxygen consumption (VO2) significantly improved from 14.2 +/- 3.7 ml/min/kg before to 21.4 +/- 5.6 ml/min/kg at a mean of 11.2 +/- 3.0 months after the transplant procedure (p < 0.001). When peak aerobic capacity was compared with a normal population, peak VO2 was < 50% of predicted in only 9 patients (12%), from 50% to 70% in 34 patients (44%), from 70% to 90% of predicted in 24 patients (31%); 10 patients (13%) were able to achieve > 90% of peak predicted VO2. The improvement seen at 6 months did not significantly change over 9 years of follow-up. Significant preoperative univariate predictors of.1-year postoperative improvement in peak VO2 were preoperative peak VO2 (p = 0.004), age (p < 0.001), ischemic cardiomyopathy (p = 0.007), and preoperative left ventricular ejection fraction (p < 0.001). Peak VO2 at 1 year in patients able to perform the test was not significantly influenced by acute rejection episodes, donor body surface area, or donor/recipient size ratio. In conclusion, exercise capacity is significantly improved within 6 months after cardiac transplantation, and maintained as long as 9 years after procedure. The magnitude of postoperative improvement is inversely related to preoperative peak VO2 and age.