The incidence of nephrotoxicity occurring with the nonionic contrast agent, iohexol, and the ionic contrast agent, meglumine/sodium diatrizoate, was compared in 1196 patients undergoing cardiac angiography in a prospective, randomized, double-blind multicenter trial. Patients were stratified into four groups: renal insufficiency (RI), diabetes mellitus (DM) both absent (N = 364); RI absent, DM present (N = 318); RI present, DM absent (N = 298); and RI and DM both present (N = 216). Serum creatinine levels were measured at -18 to 24, 0, and 24, 48, and 72 hours following contrast administration. Prophylactic hydration was administered pre- and post-angiography. Acute nephrotoxicity (increase in serum creatinine of > or = 1 mg/dl 48 to 72 hours post-contrast) was observed in 42 (7%) patients receiving diatrizoate compared to 19 (3%) patients receiving iohexol, P < 0.002. Differences in nephrotoxicity between the two contrast groups were confined to patients with RI alone or combined with DM. In a multivariate analysis, baseline serum creatinine, male gender, DM, volume of contrast agent, and RI were independently related to the risk of nephrotoxicity. Patients with RI receiving diatrizoate were 3.3 times as likely to develop acute nephrotoxicity compared to those receiving iohexol. Clinically severe adverse renal events were uncommon (N = 15) and did not differ in incidence between contrast groups (iohexol N = 6; diatrizoate N = 9). In conclusion, in patients undergoing cardiac angiography, only those with pre-existing RI alone or combined with DM are at higher risk for acute contrast nephrotoxicity.(ABSTRACT TRUNCATED AT 250 WORDS)