Persistent ovarian cancer was found in 65 out of 109 (59.6%) patients, including 17 of 31 (54.8%) evaluated by laparoscopy, 43 of 70 (61.4%) evaluated by laparotomy, and 5 of 8 (62.5%) evaluated by both procedures. The mean number of biopsies taken during laparoscopy was 9.5, compared with 15.9 during laparotomy.8 of the 39 patients who were initially evaluated by laparoscopy were converted to laparotomy. The indication for conversion to laparotomy in all 8 patients was the inability to identify gross peritoneal disease at laparoscopy. Significantly lower mean blood loss (27 versus 208 mL) was noted in patients undergoing laparoscopy compared with laparotomy (p<0.01). In addition, the mean operating time for second look laparoscopy (129 minutes) and mean hospital stay (1.6 days) were shorter than those for laparotomy (153 minutes and 6.8 days, p<0.01 for both comparisons). Recurrence after negative second-look surgery was noted in 2 out of 14 (14.3%) patients evaluated by laparoscopy and 4 out of 27 (14.8%) patients evaluated by laparotomy. Disease free survival group differences for such events had a p value = 0.62. Of the 8 patients converted to laparotomy, 1 patient would have been missed by laparoscopy. The complication rate was 27% with laparotomy and 0% with laparoscopy.