The plasma pentosidine levels in patients with renal disease were measured by a simple method which was established for plasma and urinary pentosidine determinations. The method, which can be completed within a few hours, involves pretreating plasma with proteolytic enzyme (pronase) and measuring the concentration of pentosidine in the sample by ELISA using antipentosidine antibodies. The prepared antibodies showed no cross-reaction with the raw materials for pentosidine synthesis or with compounds having similar structures. SDS-PAGE indicated that the antibodies had a high purity. The reaction of the antibodies and keyhole limpet hemocyanin-pentosidine in the competitive ELISA system was inhibited by free pentosidine. Excellent standard curves for pentosidine determination were obtained. In actual measurements of clinical samples from patients, a good correlation (r = 0.9356) was obtained between the values measured by ELISA and HPLC. The plasma pentosidine level in patients with renal disease correlated significantly with plasma creatinine, urea nitrogen, β<sub>2</sub>-microglobulin, and creatinine clearance, indicating its usefulness in evaluating the severity of renal disease. A significant elevation in plasma pentosidine levels was observed in mild renal dysfunction, whereas no significant increases in creatinine and urea nitrogen levels were detected, suggesting that the plasma pentosidine level is useful in the early diagnosis of beginning renal failure. In patients with chronic renal failure, no difference in plasma pentosidine levels was observed between diabetic nephropathy and chronic glomerulonephritis, while a significant correlation was observed with phosphatidylcholine hydroperoxide, suggesting the possibility that the plasma pentosidine level reflects injury due to oxidation. From these results, the quantitative measurement method developed by us is judged to be a superior innovation for measuring pentosidine in body fluids. The plasma pentosidine level may be useful for the early diagnosis of mild renal failure and to estimate the degree of the severity of renal diseases.