Contrast agents used for cardiac angiography are different in regard to ionicity, osmolality and physiologic effects. The nonionic contrast media have been shown to have less toxic effects and a better safety profile than do higher osmolar agents. To better assess this risk, clinically stable patients undergoing cardiac angiography were stratified according to the presence of diabetes mellitus, and level of serum creatinine, and then randomized to receive either iohexol (Omnipaque 350) or sodium meglumine diatrizoate (Renografin 76). All adverse events that occurred during and immediately after angiography were tabulated. A multivariate model was used to identify patients at increased risk for adverse outcome. The 1,390 patients were randomized to iohexol (n = 696) or diatrizoate (n = 694). Significant differences were found in the number of patients with contrast media-related adverse (iohexol vs diatrizoate: 10.2 vs 31.6%; p < 0.001) and cardiac adverse (7.2 vs 24.5%; p < 0.001) events. Severe reactions and the need for treatment were more frequent with diatrizoate than with iohexol, but there was no difference in the incidence of death. The presence of New York Heart Association classification 3 or 4 and serum creatinine > or = 1.5 mg/dl predicted a higher incidence of adverse events as a result of contrast media alone. Use of iohexol is associated with a lower incidence of all types of adverse events during cardiac angiography than is diatrizoate.