In this prospective s!Udy, prior to cholecystectomy, the diameter of the common hepatic duct was measured; duct size was then compared with probability of finding stones at operation. Of 115 patients entering the study, 36 had stones removed from the common duct at the time of cholecystectomy but only three (8°/.,) were demonstrated by ultrasonography. No stones were found in ducts""' 3 mm in size (31% patients). Only two of 26 patients with ducts measuring 4 mm had stones. As duct size increased, so did the probability of stones and all patients with ducts ;;. 9 mm in diameter had stones. It is concluded that pre-operative ultrasound provides a reliable basis for a policy of selective cholangiography.