A prospective randomized trial was undertaken to determine if selective peroperative cholangiography resulted in greater morbidity and mortality from missed common bile duct (CBD) stones. Five hundred and thirty-nine consecutive cholecystectomies were performed over a 3-year period. Two hundred and fifty-four had indications for mandatory peroperative cholangiography and were excluded from the trial. The remaining 285 patients, without a history of jaundice, pancreatitis or abnormal liver function tests, were randomized blindly into two groups. Group 1 underwent peroperative cholangiography (PC) and group 2 did not. If the surgeon found a dilated CBD at surgery then these patients were also excluded from the trial. Selective peroperative cholangiography revealed an unsuspected CBD calculus in 16 of the 132 patients (12%). Up to the time of review no patient from group 2 presented with symptoms or complications from retained CBD stones. One patient in group 1 had endoscopic removal of a retained CBD calculus 16 months after cholecystectomy. All patients were sent a questionnaire at least three years after surgery and 210 responded (74%). One hundred and thirty (62%) of the respondents had peroperative cholangiography. There were 11 deaths from unrelated causes. No difference between the two groups was found for postoperative dietary habit, dyspepsia, pain, flatulence, diarrhoea or signs of biliary obstruction. It seems from these results that a policy of selective cholangiography in our hands may miss a 12% incidence of unsuspected stones but, importantly, this does not appear to influence postoperative morbidity or mortality.(ABSTRACT TRUNCATED AT 250 WORDS)