Two hundred and seventy patients with febrile splenomegaly and coming from areas where visceral leishmaniasis (VL; kala-azar) is endemic and in whom the diagnosis of kala-azar was strongly suspected were randomly divided into 3 groups, subjected to splenic aspiration by the intercostal route, splenic aspiration by the abdominal route, and bone marrow aspiration, respectively, for demonstration of amastigotes. Pain immediately after aspiration, requiring a few analgesic tablets, occurred in 8, 16 and 20 patients of the 3 groups, respectively. Pain on the day after aspiration was reported by 2,4 and 6 patients, respectively. One patient, in the abdominal group, developed gastro-intestinal haemorrhage but this was managed with blood transfusion; he was not suffering fromVL. No patient died. The abdominal route of splenic aspiration was not feasible in 12 patients (13%), as their spleen size was less than 3 cm, and they were subsequently aspirated by the intercostal route. On the first aspiration, amastigotes were seen in 68 patients (76%) in the intercostal group, 64 patients (71%) in the abdominal group and 42 patients (46%) in the bone marrow group. Two weeks later, 15 patients (17%) in the bone marrow group and in whom amastigotes had not been detected, but whose fever continued, were subjected to intercostal splenic aspiration and amastigotes were detected. After 2 months, 3 patients in the intercostal group and 4 patients in the abdominal group gave positive aspirates, and 2 patients in the intercostal group, one in the abdominal group, and one in the bone marrow group did so at the third aspiration. Thus only 200 (74%) of 270 patients were suffering fromVL. It was concluded that, with some precautions, splenic aspiration is a safe and easy method for the diagnosis of& and the intercostal route is preferred because it is feasible in a larger proportion of cases and anatomically safer than the abdominal route, and it gives a positive result more often than bone marrow aspiration.