In 62 consecutive explorations of the common bile duct during a 2-year period, a Storz rigid choledochoscope was used to visualize the lumen of the duct before its closure over a T tube. All patients had radiological abnormalities on cholangiography: 45 had stones and 17 apparent strictures. In those with stones, the instrument was introduced after apparent removal of the stones to confirm the completeness of the manoeuvre. This examination showed that 11 patients (25 per cent) still had stones present and needed further manipulation. When the duct was seen to be clear, a T tube was inserted and the validity of the choledochoscopic findings checked by postoperative T-tube cholangiography. This confirmed the visual accuracy in all but one patient, an accuracy of 98 per cent. Of the 17 patients explored for other radiological abnormalities, choledochoscopy correctly determined the aetiology or excluded abnormalities in all cases. No complications of the procedure were encountered.