The use of standard electrocardiographic monitoring to detect cardiac allograft rejection has become unreliable since the advent of cyclosporine immunosuppression. Unipolar peak-to-peak amplitude analysis has been shown to be a quantitative measure of ischemic myocardial injury. This study was performed to determine if unipolar peak-to-peak amplitude analysis could accurately detect cardiac allograft rejection as determined by blinded endomyocardial biopsies. Ten adult mongrel dogs underwent heterotopic (n = 7) or orthotopic (n = 3) cardiac transplantation with placement of sutureless screw-in electrodes (Medtronic, Inc, Minneapolis, MN) on the anterior and posterior aspect of each ventricle. Postoperatively, animals were immunosuppressed for seven to ten days with cyclosporine and prednisone and then allowed to reject the transplant. Digitally processed intramyocardial electrograms were obtained daily. Endomyocardial biopsy was performed 1 week postoperatively and then at three to five day intervals for histological correlation. A unipolar peak-to-peak amplitude decline of 15% or greater occurred one to three days before the biopsy detection in 10 of 10 episodes of rejection. There were no false negatives and one false positive (although a small focal lymphocytic infiltrate was present). Thus, noninvasive unipolar peak-to-peak amplitude analysis was 100% sensitive and 90% specific in predicting and detecting cardiac allograft rejection.