Clinicians in Madrid have been observing and treating HIV-positive patients with visceral leishmaniasis (VL) for over a decade. As their records cover some of the co-infection cases that occurred before and after highly active antiretroviral therapy (HAART) was introduced into Spain, retrospective analysis of the records has allowed some of the effects of HAART on local VL to be determined. Encouragingly, HAART appears to have decreased the annual incidence of VL among local AIDS cases, from 4.81 cases/100 to just 0.8 case/100 (P <0.0005), a first episode of VL now appearing only when there is obvious HAART failure. Unfortunately, it does not seem to be very good at preventing VL relapses; within 24 months of antileishmanial treatment, 70% of patients who were receiving HAART had such relapses. The mean time between antileishmanial treatment and VL relapse was, however, longer when HAART was used than when it was not (20 v. 13 months). In those receiving HAART, relapses of the VL often occurred despite increasing CD4+ cell counts and undetectable HIV loads, indicating that successful treatment of the viral infection is insufficient to prevent the relapse of the leishmaniasis. These results are in general agreement with other observations made in Spain. VL relapses are possible and even frequent in HIV-positives who have no more than 200 CD4+ cells/microl, but secondary prophylaxis to prevent VL relapses may be safely suspended if a CD4+ count of >200 cells/microl can be maintained using HAART. VL also seems to hamper the immunological recovery of the HIV-positive, although HAART appears to have little effect on the clinical manifestations of VL.