Histoplasmosis is a serious opportunistic infection in patients with AIDS, often representing the first manifestation of the syndrome. Most infections occurring within the endemic region are caused by exogenous exposure, while those occurring in nonendemic areas may represent endogenous reactivation of latent foci of infection or exogenous exposure to microfoci located within those nonendemic regions. However, prospective investigations are needed to prove the mode of acquisition. The infection usually begins in the lungs even though the chest roentgenogram may be normal. Clinical findings are nonspecific; most patients present with symptoms of fever and weight loss of at least 1 month's duration. When untreated, many cases eventually develop severe clinical manifestations resembling septicemia. Chest roentgenograms, when abnormal, show interstitial or reticulonodular infiltrates. Many cases have been initially misdiagnosed as disseminated mycobacterial infection or Pneumocystis carinii pneumonia. Patients are often concurrently infected with other opportunistic pathogens, supporting the need for a careful search for co-infections. Useful diagnostic tests include serologic tests for anti-H. capsulatum antibodies and HPA, silver stains of tissue sections or body fluids, and cultures using fungal media from blood, bone marrow, bronchoalveolar lavage fluid, and other tissues or body fluids suspected to be infected on clinical grounds. Treatment with amphotericin B is highly effective, reversing the clinical manifestations of infection in at least 80% of cases. However, nearly all patients relapse within 1 year after completing courses of amphotericin B of 35 mg/kg or more, supporting the use of maintenance treatment to prevent recurrence. Relapse rates are lower (9 to 19%) in patients receiving maintenance therapy with amphotericin B given at doses of about 50 mg weekly or biweekly than with ketoconazole (50-60%), but controlled trials comparing different maintenance regimens have not been conducted. Until results of such trials become available, our current approach is to administer an induction phase of 15 mg/kg of amphotericin B given over 4 to 6 weeks, followed by maintenance therapy with 50 to 100 mg of amphotericin B given once or twice weekly, or biweekly. If results of a prospective National Institutes of Allergy and Infectious Disease study of itraconazole maintenance therapy document its effectiveness, alternatives to amphotericin B may be reasonable.