Laparoscopy and laparoscopic ultrasound are used widely in cancer staging and are perceived to prevent unnecessary open exploration in many patients. The aim of this study was to analyze the impact of staging laparoscopy in improving resectability in patients with primary and secondary hepatobiliary malignancies. Over a 10-month period (November 1, 1997 to August 31, 1998), 186 patients with primary and secondary hepatobiliary cancers were submitted to operation for potentially curative resection. One hundred four patients staged laparoscopically (LAP) before laparotomy were compared prospectively to 82 patients undergoing exploration without laparoscopy (NO LAP). Assignment to each group was not random but was based on surgeon practice. Demographic data, diagnoses, the extent of preoperative evaluation, and the percentage of patients resected were similar in the two groups. Laparoscopy identified 26 (67%) of 39 patients with unresectable disease. In the NO LAP group, 28 patients (34%) had unresectable disease discovered at laparotomy. In patients with unresectable disease and submitted to biopsy only, the operating times were similar in the two groups (LAP 83 +/- 22 minutes vs. NO LAP 91 +/- 33 minutes; P = 0.4). However, laparoscopic staging significantly reduced the length of hospital stay (LAP 2.2 +/- 2 days vs. NO LAP 8.5 +/- 8.6 days; P = 0.006). Likewise, total hospital charges, normalized to 100 in the NO LAP patients, were significantly lower in the LAP group (LAP 54 +/- 42 vs. NO LAP 100 +/- 84; P = 0.02). Staging laparoscopy identified the majority of patients with unresectable hepatobiliary malignancies, significantly improved resectability, and reduced the number of days in the hospital and the total charges. The yield of laparoscopy was greatest for detecting peritoneal metastases (9 of 10), additional hepatic tumors (10 of 12), and unsuspected advanced cirrhosis (5 of 5) but often failed to identify nonresectability because of lymph node metastases, vascular involvement, or extensive biliary involvement. Eighty-three percent of patients subjected to laparotomy after laparoscopy underwent a potentially curative resection compared to 66% of those who were not staged laparoscopically.