Drainage after cholecystectomy remains routine despite the lack of scientific supportive data. Numerous clinical studies in the past have attempted to address this controversy but have failed to resolve the issue for different reasons. These include retrospective design, inclusion of only selected cases, and randomization before surgery. In this study 479 patients undergoing cholecystectomy were randomly allocated to a drainage group (a high-pressure suction drain in Morison's pouch for 48 hours) or a nondrainage group. Randomization was performed at the time of peritoneal closure. All patients undergoing cholecystectomy, both elective and urgent, were included and the operations were performed by all grades of surgeons. There were two deaths from cardiopulmonary causes, both in the drainage group. No patient required reoperation in either group. The incidence of both wound infections (15 vs 5; p less than 0.05) and chest infections (56 vs 19, p less than 0.02) was significantly higher in the drainage group. Three hundred fifty-six patients underwent abdominal ultrasonography 72 hours after surgery. The number of subhepatic fluid collections thus detected was significantly higher in the patients who received a drain (17 vs 6, p less than 0.05). None of these collections was clinically significant. The postoperative hospital stay was longer in the patients with drains (10.3 vs 9.1 days), but this difference failed to reach statistical significance. We conclude from this study that the use of a drain after cholecystectomy serves no useful purpose and is potentially harmful. This practice should be abandoned.