Since the 1940s meglumine antimoniate (MA) has been the only first-line drug for visceral leishmaniasis (VL) treatment in Italy. From 1991 through 1994, several patients of all ages, representing 1/3 of all immunocompetent VL patients reported during that period, were enrolled in clinical trials of liposomal amphotericin B (L-AmB), which led to a novel, safe, short course of VL treatment as an alternative to MA. In the same period, other lipid-associated AmB drugs were registered in Italy for the treatment of fungal infections, i.e., AmB colloidal dispersion (ABCD) and AmB lipid complex (ABLC). A retrospective analysis was performed on data collected at the Unit of Protozoology of Istituto Superiore di Sanità, Rome, to assess whether changes have occurred in first-line drug regimens adopted in Italy for routine VL treatment, during the 1995-2002 period. The sample consisted of immunocompetent individuals clinically suspected for VL, in whom the disease was confirmed by the examination of serum and bone marrow specimens sent to the Unit by hospitals from throughout the country. Relevant information on patients was then recorded, which included drug regimens used and post-therapy results. We recorded treatment information for 630 patients, representing a large proportion (55.5%) of 1,135 immunocompetent individuals with VL reported in Italy from 1995 through 2002. About half were children (306). Every year, patients were referred by 19 to 42 hospitals, with a range of 1 to 30 patients per hospital. MA was the first-line drug used in 159 patients (25.2%). However, the proportion of MA-treated patients has steadily decreased from 55.9% in 1995 to 1.0% in 2002. We recorded the failure of MA therapy in 16 patients (10.1%), who were successfully retreated with a L-AmB regimen. The rate of MA failures significantly increased in recent years, from 5.3% in 1995 to 36.4% in 2000 (p = 0.01). AmB drugs have been the only alternative drugs used in the remaining 471 patients (74.8%). L-AmB accounted for most regimens (441, 93.6%). The proportion of patients treated with any AmB-based drugs increased from 44.1% in 1995 to 99.0% in 2002. Drug treatment was unsuccessful in 15 patients (3.2%), who were successfully retreated with a high-dose L-AmB regimen. This rate was significantly lower than the MA failure rate (p = 0.001). Results have shown a countrywide change in therapy over the period considered. A traditionally effective, but moderately toxic drug (MA) has been almost fully replaced by a new compound (L-AmB) with negligible toxicity, in an epidemiologic context of disease reemergence. Furthermore, short courses of 6 to 7 days, as required for lipid-associated AmB, are highly cost-effective if compared with 21- to 28-day courses needed for standard MA treatment.