Laparoscopy was evaluated in thirty-seven patients from a group of 132 consecutive patients who were treated for blunt or penetrating injury to the abdomen. A total of twenty-three patients underwent laparoscopy and laparotomy. The findings at laparotomy correlated with laparoscopy. Fourteen patients underwent laparoscopy only, and there were no proved false-negative results. Of the 132 patients considered for laparotomy, 118 underwent abdominal exploration. Laparotomy was considered unnecessary in twenty-five of the 118 patients (21 per cent) and in retrospect, laparoscopy could have identified in each patient the presence of a minor injury or no injury at all. Laparoscopy is a useful method for evaluating blunt and penetrating injuries to the abdomen in selected patients.