The prospective study involved 50 patients without known renal masses or previous renal surgery who received intravascular contrast medium for various radiologic procedures and subsequently underwent unenhanced computed tomography (CT) of the kidneys 22-26 hours later. In 42 patients, cortical readings did not exceed 50 HU and corticomedullary attenuation differences (CMADs) did not exceed 10 HU. In seven patients (intermediate group), there was persistent cortical enhancement (mean attenuation, 58.3-84.8 HU; CMAD, 10.3-40.3 HU), but there was no significant rise in levels of blood urea nitrogen or serum creatinine. A single patient had contrast nephropathy and exhibited a mean cortical attenuation of 141.6 HU and a mean CMAD of 85.8 HU on the 24-hour CT scan. The authors have seen two additional cases of contrast nephropathy presaged by mean cortical attenuations of 162 and 209 HU on the 24-hour CT scan. One of these patients had been in the intermediate group in the prospective study. A cortical attenuation of 55-110 HU on a 24-hour postcontrast scan may help identify a group of patients with subclinical renal impairment who are at high risk of nephropathy on subsequent exposure to contrast medium, while cortical attenuations in excess of 140 HU seem to be an early indicator of contrast nephropathy.