In patients who have undergone successful orthotopic heart transplantation, a noninvasive method for rejection surveillance would reduce the frequency of endomyocardial biopsy, guide the timing of biopsies, and allow for more frequent monitoring. This study identified the pattern of change in echocardiographically determined indexes of left ventricular mass, volume, and function that characterized biopsy-positive acute rejection in adult heart transplant patients receiving triple-drug immunosuppressive therapy and describes a simple computer-driven algorithm capable of identifying rejection with high sensitivity and specificity. Two-dimensional and M-mode echocardiography and Doppler color flow analyses were performed within 24 hours of endomyocardial biopsy. M-mode echocardiograms of the left ventricle were digitized and analyzed blinded to the biopsy interpretation, using a computer-assisted measurement format, for size, mass, and wall motion in systole and diastole. Twenty-nine studies were retrospectively analyzed to define the echocardiographic pattern characteristic for rejection. Left ventricular chamber size decreased, and indexes of diastolic function were significantly depressed in patients with biopsy evidence of moderate or severe rejection. However, no single parameter was sufficiently sensitive to detect all episodes of rejection partly because of differences between patients in the echocardiographic manifestations of acute rejection. To accommodate this patient variability, multiple echocardiographic parameters were clustered into a unique scoring algorithm (ECHO score). When applied prospectively to 49 studies, the likelihood that a patient would have an ECHO score not indicative of rejection but with moderate/severe rejection on biopsy was low (less than 3%) or a negative predictive value of 97.4%.(ABSTRACT TRUNCATED AT 250 WORDS)