Preexisting renal impairment is an all-encompassing risk factor for radiocontrast-associated nephrotoxicity. Renal impairment appears to be associated with the inadequate production of renal prostaglandins at the critical time of radiocontrast administration and for a variable time period afterward. We prospectively studied 130 patients with chronic renal insufficiency (serum creatinine > or =1.5 mg/dL) who were undergoing radiocontrast administration. Using a double-blind, randomized, prospective technique, patients were assigned to either placebo or one of three prostaglandin E1 (PGE1) treatment groups (10, 20, or 40 ng/kg/min). Infusion was started 60 +/- 30 minutes before the administration of radiocontrast and was continued for a total of 6 hours. In the placebo group, radiocontrast administration resulted in a mean increase (+/- SD) in serum creatinine of 0.72 +/- 1.15 mg/dL at 48 hours. This increase was less in each of the PGE1 treatment groups after 48 hours, with a significant difference between placebo and the 20 ng/kg/min PGE1 group (P = 0.01). Using baseline adjusted means, analysis of covariance with baseline serum creatinine as the covariable demonstrated significant differences between the placebo and 20 ng/kg/min PGE1 group (P = 0.03) and between the placebo and 10 ng/kg/min PGE1 group P = 0.047). In a subgroup analysis of the diabetic patients, the increase in serum creatinine was less pronounced in the three PGE1 groups versus the placebo group, and the 20 ng/kg/min PGE1 group had the most favorable outcome. The parenteral administration of PGE1 immediately before radiocontrast exposure and continued for a period of 5 to 5.5 hours significantly reduced the elevation of serum creatinine poststudy. The most effective of the three PGE1 dosing regimens was 20 ng/kg/min.