As acute tubular necrosis (ATN) is still an important cause for postoperative malfunction of renal grafts, it would be useful to have a method predicting such a complication. We investigated the possibility to predict ATN by measuring the ratio of phosphomonoesters (PME, largely consisting of adenosine monophosphate) and inorganic phosphate (Pi) in the renal tissue, using P-31 magnetic resonance spectroscopy (MRS) during the cold ischemia period. Assuming that this ratio reflects the tissue high-energy phosphate status, we studied five kidneys from living related donors (LRD), 28 kidneys from heart beating donors (HBD) and nine kidneys from non-heart beating donors (non-HBD). All kidneys were preserved with a phosphate free solution. We found an inverse relation between the time of P-31 MRS and the PME/Pi ratio, suggesting a graded decay of tissue high energy phosphates during cold ischemia. The PME/Pi ratio was highest in grafts from LRD (2.65 +/- 0.50, no ATN), intermediate in grafts from HBD (1.65 +/- 0.41, 21% ATN) and lowest in those derived from non-HBD (1.05 +/- 0.47, 56% ATN). The differences in PME/Pi ratio between the groups was statistically significant (P < 0.01). Moreover, the ratio was significantly lower in grafts developing ATN (1.73 +/- 0.41 vs. 1.35 +/- 0.29 in the HBD group, 1.41 +/- 0.24 vs. 0.76 +/- 0.36 in the non-HBD group, P < 0.05). These observations point to a general relation between the pre-transplant kidney PME/Pi ratio and the development of ATN. However, the predictive value of a low PME/Pi ratio was too low (36%) to reliably predict development of ATN in individual cases. Extrapolation of the PME/Pi ratio to the time of nephrectomy increased the predictive value to 67%.