The rapid introduction of laparoscopic cholecystectomy has been associated with an apparently increased incidence of bile duct injury which has provoked worldwide concern. The true incidence and mechanism of iatrogenic ductal injury during the development of this procedure remain unclear. To assess this, the introduction of laparoscopic cholecystectomy in the West of Scotland has been audited prospectively over a 5-year period. All cases of biliary ductal injury have been independently reviewed. Some 48 surgeons undertaking laparoscopic cholecystectomy in 19 hospitals submitted prospective data between September 1990 and September 1995. A total of 5913 laparoscopic cholecystectomies were attempted with 98·3 per cent completion of data collection. During this period 37 laparoscopic bile duct injuries occurred. The annual incidence peaked at 0·8 per cent and has fallen to 0·4 per cent in the final year of audit. Injuries occurred after a median personal experience of 51 (range 3–247) laparoscopic chole cystectomies in 22 surgeons. Major bile duct injuries occurred in 20 of 37 patients, giving an incidence of 0·3 per cent. Five mechanisms for laparoscopic ductal injury were identified, including tenting, confluence and diathermy injuries as well as the classical and variant classical types. Ductal injuries were discovered at operation in 18 patients with consequent repair giving a good clinical outcome in 17. Contributory factors (severe inflammation, aberrant anatomy and poor visualization) were present in only 13 of 37 cases. This audit suggests that, at least in the introductory period, laparoscopic cholecystectomy is associated with an overall bile duct injury rate higher than that reported previously after open cholecystectomy, although the incidence of major ductal injury is similar. The late downward trend in bile duct injury, however, suggests there may be a prolonged learning curve for this procedure. Improved understanding of the mechanism of injury may lead to yet further reductions in this complication.