Rats with untreated diabetes mellitus are protected from gentamicin-induced nephrotoxicity. In order to evaluate the role of hyperglycemia, glycosuria, and polyuria in this phenomenon, miniosmotic pumps filled with insulin were implanted for 15 days in seven female Sprague-Dawley rats with streptozotocin-induced diabetes mellitus. Plasma glucose levels were successfully maintained under 126 mg/dl. To serve as the control group, eight age-matched diabetic (plasma glucose > 400 mg/dl) rats had miniosmotic pumps placed delivering only Ringer's solution. Six days after placement of the pumps, gentamicin (40 mg/Kg/day) was administered to all animals for 9 days. The insulin-treated diabetic rats exhibited clear signs of nephrotoxicity by Day 6 of gentamicin, whereas the diabetic control group remained free from any functional or morphological evidence of proximal tubular damage throughout the 9 days of the aminoglycoside administration. At the end of the experiment, the creatinine clearance in the insulin-treated diabetic group was 45% lower than in the untreated diabetic group (P < 0.005). In addition, there was a rise in plasma creatinine (P < 0.02), muramidase appeared in the urine, and mild patchy acute tubular necrosis of the renal cortex was observed by light microscopic examination. The insulin-treated group also accumulated more gentamicin in the renal cortex than the untreated animals (P < 0.005). It is concluded that protection against the nephrotoxic effects of gentamicin is a feature of untreated experimental diabetes mellitus in the rat and that correction of the hyperglycemic state with insulin reverses this resistance.