We evaluated the role of 31P MR spectroscopy in the diagnosis of renal transplant allograft dysfunction. Thirty-six 31P MR spectroscopy examinations were prospectively performed in 35 patients with renal allografts. The study was performed in two phases. In the first phase, 12 transplant recipients with normal graft function were studied as normal controls. During phase two, 24 31P MR spectroscopy studies were performed in patients at the time of renal transplant biopsy for allograft dysfunction. Twenty-one of these studies were technically adequate. Pathologic analysis of the biopsy specimens showed evidence of allograft rejection in 14 and no rejection in seven. Various phosphorus metabolite ratios were calculated for each patient, including phosphodiesters/phosphomonoesters (PDE/PME), phosphomonoesters/inorganic phosphate (PME/Pi), and inorganic phosphate/adenosine triphosphate (Pi/ATP). The PDE/PME and Pi/ATP ratios in the allografts with rejection differed significantly from the corresponding metabolite ratios in patients without rejection (p = .017 and p = .024, respectively). A PDE/PME ratio exceeding 0.8 had a sensitivity of 100% and specificity of 86% for predicting rejection. A Pi/ATP ratio greater than 0.6 had a sensitivity of 72% and a specificity of 86% for predicting rejection. We conclude that 31P MR spectroscopy may be useful as a noninvasive method for evaluating renal metabolism during episodes of transplant allograft dysfunction.