OBJECTIVE--To evaluate the prevalence of cardiac toxoplasmosis in a series of 182 necropsies performed between 1987 and 1991 on patients infected with the human immunodeficiency virus (HIV), to correlate this prevalence with the ante mortem diagnosis of cardiac involvement, and to assess the role of such cardiac lesions in the immediate cause of death. PATIENTS AND METHODS--Complete necropsies of 182 HIV-infected patients (48 women, 134 men) were performed consecutively between 1987 and 1991. Risk factors, identified in 174 cases, included drug abuse (111/182), homosexuality (51/182), and blood transfusions (12/182). 16 samples were systematically obtained from each heart for histological study. If trophozoites or lymphocytic myocarditis were seen, immunohistochemical investigations were carried out with polyclonal antibodies for Toxoplasma gondii. An ultrastructural study was performed in four patients with toxoplasma myocarditis. Myocardial lesions were defined by the Dallas classification. Clinical data (and information on electrocardiograms and echocardiograms) were obtained from medical records. RESULTS--Cardiac toxoplasmosis was diagnosed at necropsy in 21 (12%) patients. Cardiac lesions were associated with toxoplasmic encephalitis in 18 patients and were solitary in three patients. Acute diffuse myocarditis was present in 6/21, rare foci of myocarditis were seen in 8/21, and intramyocytic toxoplasmic cysts without any inflammatory reaction or necrosis were seen in 4/21. Anti-toxoplasma immunolabelling showed cardiac toxoplasmosis in three patients with lymphocytic myocarditis. Particles with the ultrastructural characteristics of Toxoplasma gondii trophozoites were seen in four cases. Six patients had presented with cardiac symptoms, confirmed by electrocardiographic and echocardiographic abnormalities during their disease course, and their cardiac lesions were directly responsible for the death. CONCLUSION--Cardiac toxoplasmosis was common in this necropsy series of HIV-infected patients. Cardiac toxoplasmosis had been suspected clinically in four patients. Myocardial lesions were generally asymptomatic and were not discovered until necropsy. Solitary cardiac involvement was not uncommon reflecting parasite reactivation at a myocardial site. The incidence of cardiac toxoplasmosis in this group of immunodepressed subjects from an area with a high prevalence of this parasitic disease justifies regular follow up of such patients by electrocardiography and echocardiography as well as immediate administration of anti-toxoplasma treatment should sudden heart failure occur.