Routine drainage of liver wounds created by trauma has recently been challenged, prompting a prospective, randomized trial of drainage via a Penrose dam versus no drain in patients having emergency laparotomy for abdominal trauma. We excluded cases in which definite bile leak was noted at operation. Of 167 patients studied, six had obligatory drainage because of obvious bile leak. Among the remaining 161 patients, there was no significant difference as to demographics, mode of injury, volume of blood lost or used for resuscitation, incidence and severity of shock, number and types of associated injuries, or magnitude of liver wound between the 78 allocated to drainage and the 83 left without a drain. Resultant mortality, duration of hospitalization, incidence of wound and/or intra-abdominal infection, and likelihood of subsequent bile fistula were not different. Such data support the routine use of a drain only if bile leakage from the liver wound is found at laparotomy. Without obvious bile leak, drainage of a specific liver injury does not appear to be necessary.