In order to compare their capacity to visualize the bile ducts, ultrasonography, computed tomography, and cholescintigraphy were performed in 56 consecutive jaundiced patients in whom extrahepatic cholestasis was clinically suspected. The predictions as to the patency of the large bill ducts were compared with the final diagnoses made on the basis of direct cholangiography together with autopsy, biopsy, operative findings, and the clinical course. Thirty-nine patients (70%) had obstructed bile ducts, and 17 (30%) had patent large bile ducts. Using a simple scoring scale with 112 points as the maximum, ultrasonography obtained 72 points, computed tomography received 56 points, and cholescintigraphy totalled 37 points. Nonsignificant trends were found in favor of ultrasonography as compared with computed tomography and of computed tomography as compared with cholescintigraphy (p greater than 0.05), whereas ultrasonography was significantly better than cholescintigraphy (p = 0.01). However, because computed tomography is expensive and may imply a higher number of secondary direct cholangiographies than ultrasonography, we recommend ultrasonography as the first choice for noninvasive bile duct visualization. Computed tomography is an alternative method, whereas cholescintigraphy cannot be recommended.