To evaluate the usefulness of phosphorus-31 magnetic resonance spectroscopy (31P-MRS) in assessing male infertility, we compared it with conventional semen analysis. Specimens were obtained from otherwise healthy patient groups as follows: group A, 7 fertile control subjects; group B, 12 azoospermic men after vasectomy; and group C, 11 patients presenting for infertility evaluation. Correlations between established semen analysis parameters and the 31P-MRS-derived ratio of glycerylphosphorylcholine to total phosphate (GPC/TP) were investigated. Group A controls had a mean GPC/TC ratio of 0.10 +/- 0.05, which was the same as that of group C. With the exception of significantly lowered motility and normal morphology in group C (p less than 0.001 and 0.05, respectively) semen analysis parameters in these two groups were similar. In contrast, the GPC/TP ratio in group B (0.05 +/- 0.04) was significantly different from the control (p less than 0.05), which appropriately reflected complete vasal occlusion. The results suggest that a significant portion of seminal GPC is derived from epididymal secretion and that 31P-MRS is useful for monitoring the GPC/TP levels when assessing epididymal function and male infertility.