Peritoneal lavage was studied prospectively in 232 patients with blunt abdominal trauma and proved to be a rapid, safe, prompt, and reliable indicator of the presence or absence of hemoperitoneum. Ninety-one percent of the significant abdominal injuries were identified at the time of initial emergency room evaluation. However, 30% of the patients with hemoperitoneum did not have significant injuries. A simple bedside method of measuring the amount of blood in the lavage fluid has allowed us to identify patients with minimal hemoperitoneum. By evaluating these patients more intensively, the remaining injuries are detected promptly, and patients with hemoperitoneum from trival injuries are spared unnecessary celiotomies. The diagnostic error was 4%, but no significant injuries were overlooked. Preoperative time, blood transfusions, and delays have been reduced, and admissions for "observation" have been eliminated.