Monitoring human cardiac allograft rejection is currently accomplished by endomyocardial biopsy. Available noninvasive methods for identifying rejection have lacked the necessary sensitivity or specificity, or both, for routine clinical application. In vivo phosphorus-31 (P-31) nuclear magnetic resonance (NMR) spectroscopy has been used for monitoring phosphorus metabolism in both animal models and humans. In the present study this technique was employed as a noninvasive means to assess the bioenergetic processes that occur during cardiac allograft rejection in a rat model. Brown Norway rat hearts were transplanted subcutaneously into the anterior region of the neck of Lewis rat recipients (allografts). Control isografts employed Lewis donors and recipients. Phosphocreatine to inorganic phosphate (PCr/Pi), phosphocreatine to beta-adenosine triphosphate (PCr/ATPβ), beta-adenosine triphosphate to inorganic phosphate (ATPβ/Pi) ratios and pH of the transplanted hearts were monitored using surface coil P-31 NMR spectroscopy (at 4.7 tesla) daily for 7 days. To allow recovery from the compromise induced by the surgical procedure, the measurements obtained on day 2 were taken as a baseline. PCr/Pi was unchanged or increased in the iso-grafts but decreased continually in allografts, with the difference becoming significant by day 4 when compared with levels in day 2 allografts (p < 0.005) and by day 3 when compared with levels in the isograft group (p < 0.05). PCr/ATPβ in isografts did not change throughout the study; however, allografts demonstrated a significant decrease as early as day 3 (p < 0.01), although a significant difference between isografts and allografts did not become manifest until day 4 (p < 0.005). The ATPβ/Pi differences between the two groups were similar to those for PCr/ATPβ, achieving a significant difference between groups on day 4 (p < 0.05). In the isograft group, a significant increase in ATPβ/Pi was present on day 3 (p < 0.05). The isograft group showed no change in intracellular pH; however, the allograft group demonstrated an initial alkaline shift followed by acidosis. This study, coupled with the development of new methods to apply P-31 spectroscopy clinically, suggests that the abnormal bioenergetics associated with cardiac allograft rejection may be a useful means for following cardiac transplant patients. © 1987, American College of Cardiology Foundation. All rights reserved.