The value of ascitic fluid adenosine deaminase activity in distinguishing tuberculosis from other causes of ascites was examined in a retrospective study of 41 patients with bacteriologically confirmed tuberculous peritonitis and 41 control patients, matched for age and sex, with ascites of other causes (12 alcoholic cirrhosis, 5 cryptogenic cirrhosis, 12 malignant disorders, 3 pancreatitis, and 9 miscellaneous causes). The mean ascites adenosine deaminase activity was 99.8 (SD 49.1) in tuberculous patients and 14.8 (8.4) U/l in control patients (p less than 0.0001). A cutoff of 32.3 U/l had a sensitivity of 95% and specificity of 98% in distinguishing between the two groups. In a subsequent prospective study of 64 patients with ascites, 11 were found to have tuberculosis. Of the others, 23 had cirrhosis (18 alcoholic, 5 cryptogenic), 17 malignant disorders, 3 pancreatitis, 5 cor pulmonale, 3 congestive cardiac failure, 1 systemic mastocytosis, and 1 renal failure and hypothyroidism. The mean ascites adenosine deaminase activity was 112.6 (45.0) U/l in the patients with tuberculous ascites and 16.3 (36.7) U/l (p less than 0.0001) in those with ascites of other causes. In this study, adenosine deaminase had a sensitivity of 100% and specificity of 96% in discriminating tuberculosis from other causes of ascites. These findings suggest that the ascitic fluid adenosine deaminase activity may be used to identify patients in whom the diagnosis of abdominal tuberculosis must be pursued.