Abdominal laparoscopy was performed on 200 patients with undiagnosed ascites. It was unsuccessful in one patient with tuberculous peritonitis because of extensive adhesions. A presumptive diagnosis of tuberculous peritonitis based on clinical findings and peritoneal tubercles or adhesions visualized during laparoscopy was made in 90 of these patients. The diagnosis was confirmed in 88 by histopathology, bacteriology, or therapeutic response. Two of the 109 remaining patients who had other presumptive diagnoses made during laparoscopy were eventually confirmed to be cases of tuberculous peritonitis. Of 91 patients with tuberculous peritonitis included in this series, 79% were females, with the majority (79%) of them being of child-bearing age. Half had been ill for longer than one month. The most frequent complaints were abdominal pain, fever, anorexia, night sweats, abdominal swelling, and weight loss. Ascites, fever, wasting, pallor, and abdominal tenderness were common findings. Ultrasonography demonstrated ascites in all patients who underwent this procedure; 21% also had adhesions. Pleural effusion was present in 15% and pulmonary tuberculosis was detected in only two patients. Biopsy samples taken during laparoscopy showed that 60% had noncaseous granulomas and 33% had caseous granulomas. Mycobacterium tuberculosis was detected in 77%, with guinea pig inoculation having the highest sensitivity, followed by culture, and lastly by acid-fast smear. Mycobacterium tuberculosis was isolated more easily from biopsy samples than from ascitic fluid. Nine of 20 M. tuberculosis isolates that were identified as to species were M. bovis. Tuberculous peritonitis, a frequent cause of febrile ascites in Egyptian women, was easily diagnosed by histopathologic and bacteriologic studies of biopsy samples taken at laparoscopy. All patients responded rapidly to antituberculosis therapy.