We evaluated the accuracy of MR imaging (MRI), mainly MR cholangiopancreatography (MRCP), in the diagnosis of choledocholithiasis in comparison with ultrasound (US), computed tomography (CT), direct cholangiography, and intravenous cholangiography (DIC). Thirty-seven patients with biliary disease diagnosed by surgery and direct cholangiography underwent MRI with T1-weighted images (T1-WI), T2-weighted images (T2-WI), and MRCP (source images and projection images). The rate of stone detection was evaluated for each MRI image and modality. Fifteen of 37 patients were found to have common bile duct stones at surgery. The depiction rate of T1-WI, T2-WI, source images, and projection images were 7%, 67%, 93%, and 53%, respectively. The depiction rate of MR, US, CT, direct cholangiography, and DIC were 100% (15/15), 25% (3/12), 64% (9/14), 71% (5/7), and 57% (4/7), respectively. In addition, there were two false-positive cases, one of duodenal diverticulum (Lemmel's syndrome) and the other of hemobilia. MRI had a sensitivity of 100%, specificity of 91%, and accuracy of 94% in the diagnosis of common bile duct stones. In conclusion, MRI is useful for evaluating suspected choledocholithiasis. In spite of the high depiction rate of the source images, other images should also be taken into consideration in the diagnosis of choledocholithiasis.