We conducted a randomized, blinded, prospective study evaluating hepatic parenchymal density changes during dynamic bolus CT (180 ml of contrast material given IV) and delayed CT (5 hr after 60 g of iodine) in order to compare the enhancement characteristics of an ionic contrast agent (iothalamate-60) and two nonionic agents (iohexol-300 and iopamidol-300). A total of 75 patients with known or suspected cancer were studied (25 patients per contrast agent). After a baseline unenhanced CT scan was obtained, dynamic bolus and delayed CT scans were obtained for all patients with one of the three contrast agents. The density of the liver and spleen was measured in Hounsfield units (H) for unenhanced CT, dynamic bolus CT, and delayed CT. The average percentage of enhancement was calculated as follows: postcontrast density minus precontrast density was divided by precontrast density and then multiplied by 100. For dynamic bolus CT, the average percentage of enhancement of the liver was 105% when iohexol-300 was used, 98% when iopamidol-300 was used, and 83% when iothalamate-60 was used. No significant difference was seen between the postcontrast enhancement of the three contrast agents on dynamic bolus CT scans (p greater than .05). For delayed CT, the average percentage of enhancement of the liver was 34% when iothalamate-60 was used, 28% when iopamidol-300 was used, and 16% when iohexol-300 was used. Both iothalamate-60 and iopamidol-300 showed superior enhancement on delayed CT, compared with iohexol-300 (p = .0001). We conclude that for dynamic bolus CT, all three contrast agents are similar, with no statistically significant differences in postcontrast enhancement of the liver. For delayed CT, however, hepatic enhancement with iothalamate-60 and iopamidol-300 is statistically superior to that with iohexol-300.