Results of percutaneous transhepatic cholangiography (PTC) with the Chiba needle were compared with those obtained by endoscopic retrograde cholangiopancreatography (ERCP). PTC was used in 102 cases, most often after ERCP had been inconclusive or had failed. Overall success rate of PTC was 79%. Contrast medium could be injected in only 66% of cases with normal-sized bile passages, but in 98% of those with enlarged passages. The most frequent side effect was temporary pain in the right upper abdominal quadrant. One patient had transitory severe peritoneal irritation with shock. Emergency operation was never required. PTC is technically simpler, takes less time and causes less stress to the patient, but it is less informative than ERCP. If, therefore, both methods are mastered technically, ERCP should be employed first. If it fails to provide the diagnosis, PTC can be performed at once. Biliojejunal anastomoses and pancreatic pseudocysts are absolute indications for the primary use of PTC to investigate the cause of biliary stasis. On the other hand, allergies to iodine or contrast medium and disorders of coagulation are absolute indications for ERCP.