We compared the diagnostic accuracy of multidetector-row computed tomography (MDCT) with multiplanar reconstruction (MPR) images to non-contrast magnetic resonance imaging (MRI) with magnetic resonance cholangiopancreatography (MRCP) (MRI/MRCP) for evaluating obstructive jaundice. MDCT and MRI/MRCP images from 53 patients with suspected obstructive jaundice were interpreted by two radiologists. These readers evaluated the images to determine level of obstruction, to differentiate between benign and malignant lesions, and to state the first-choice diagnosis with degree of confidence. We analyzed the obstruction levels in 50 patients excluding 3 patients who did not undergo direct cholangiography (DC). Both MDCT and MRI/MRCP showed almost perfect agreement with DC in two readers (statistic weighted kappa≥0.80) in the determination of obstruction level. The mean area under the receiver operating characteristic curve for differentiating benign from malignant lesions was significantly (p=0.02) larger in MDCT (0.98) than in MRI/MRCP (0.86). We analyzed the first-choice diagnoses for 39 patients excluding 14 patients without final diagnosis confirmed. Readers had, out of 78 interpretations, a high confidence level in their first-choice diagnoses for 44 (56%) and 23 (29%) interpretations using MDCT and MRI/MRCP, respectively. In the interpretations made with high confidence level, 98% (43/44) and 91% (21/23) were correct for MDCT and MRI/MRCP, respectively. In conclusion, MDCT with MPR images is as accurate as MRI/MRCP for evaluating the biliary duct obstruction level, and has high diagnostic accuracy in evaluating the cause of jaundice. MDCT can provide sufficient information on the level of biliary obstruction and cause of obstructive jaundice.