Preoperative abdominopelvic computed tomography results and operative findings were compared in 52 patients undergoing second-look laparotomy to confirm tumor status. Seventeen true-negative, 22 false-negative, and 13 true-positive scans were found. The sensitivity was 0.38, specificity was 1.0, and diagnostic accuracy was 0.58. Negative studies were associated with positive findings at laparotomy in 42% of all cases. Fourteen patients were identified who had computed tomography that would have enabled an attempt at the diagnosis of persistent cancer by computed tomography-directed needle aspiration or biopsy, thus avoiding laparotomy. Assuming 80% accuracy of needle aspiration, the cost of computed tomography in all 52 patients is considerably outweighed by the savings that could have been realized by eliminating the need for second-look surgery in these 11 women.