Present treatment strategies for kala-azar (visceral leishmaniasis, VL) include use of first line drug sodium antimony gluconate (SAG) to all patients but a large number of patients do not get relief with this drug. If a patient does not respond to a full course of SAG, a second or third line drug is given. We undertook this study to test whether an improved outcome can be achieved by employing a strategy of treatment based on culture and sensitivity of amastigotes to SAG compared with conventional empirical treatment.In a double-blind, randomized, controlled trial done in Balaji Utthan Sansthan, Patna, of the 181 patients screened,140 were finally randomly allocated to two groups A and B; group A patients were treated with SAG if their amastigotes were sensitive to SAG, and all patients in group B were treated with SAG to start with. Primary outcome measured was as no relapse within 6 months of follow up after cure and other outcomes measured were period of stay of patients in hospital, expenditure involved in the treatment, and infectivity periods of two groups, two-third of treatment period and whole of untreated period were taken as infectivity period. SAG was used at a dosage of 20 mg/kg given deep intramuscular injections in buttock for 28 days, amphotericin B (AMB) given at a dose of 1 mg/kg body wt daily for 20 days as a slow intravenous infusion in 5 per cent dextrose.Of the 70 patients in group A, 29 patients whose amastigotes were sensitive to SAG were treated with SAG, 2 patients were withdrawn due to drug toxicity; and 2 relapsed within 6 months of follow up and ultimate cure occurred in 25 (86.2%) patients only. Of the 70 patients in group B treated with SAG, 5 (7.1%) patients withdrew due to drug toxicity, 35 patients (50%) did not respond to treatment, 5 (7.1%) relapsed during 6 months of follow up and thus only 25 patients (35.7%) were ultimately cured. The difference between the two groups was significant (P<0.001). No patient died during treatment due to any toxicity because of early withdrawal of patients from treatment apprehending toxicity. Patients whose amastigotes were resistant to SAG, withdrawn from the study due to SAG toxicity, relapsed after cure with SAG, and who did not respond to SAG in both the groups were treated with AMB and all were cured. Groups B and A patients spent 3065 and 2340 days respectively in hospital, group B 1.3 times more than group A. The likely period of spread of parasites in society was 1965 days in group B and 1644 days in group A, group B 1.4 times more than group A. The total expenditure on treatment in groups B and A was dollars 65,575 and dollars 50,590 respectively; group B patient had to spend 1.3 times more than group A.A new strategy for treatment of kala-azar based on culture and sensitivity of amastigotes improved the cure rate, saved expenditure on the patient's treatment, patients had to stay for shorter periods in hospital and reduced the chance of spread of SAG resistant disease in society. Till the government opts for better drugs, the treatment based on culture and sensitivity of the parasites to SAG may be a better method.