To determine the value of aerosol amphotericin B inhalations for prevention of invasive pulmonary aspergillosis (IPA), we initiated a prospective randomized multicenter trial. The scheduled intent-to-treat interim analysis included 115 patients (30%) with prolonged neutropenia after chemotherapy for acute myeloid leukemia, acute lymphoblastic leukemia/high-grade non-Hodgkin's lymphoma, or solid tumors undergoing autologous stem cell transplantation. Sixty-five patients had been randomized to receive prophylactic aerosol amphotericin B inhalations at a dose of 10 mg twice daily (group A); for the remaining 50 patients no aerosol amphotericin B prophylaxis was used (group B). No serious side effects from amphotericin B inhalations occurred, but coughing (54%), bad taste (51%), and nausea (37%) caused early cessation of aerosol amphotericin B prophylaxis in 23% (15/65) of courses. In group A, the incidence of proven, probably, or possible IPA was 5% (3/65) as compared with 12% (6/50) in group B (p > 0.05). Microbiologically documented bacterial pneumonias were observed in 5/65 (8%) patients in group A and in 1/50 (2%) patients in group B (p > 0.05). Thus, no reduction in incidence of IPA from use of prophylactic aerosol amphotericin B inhalations was found in this interim analysis. As there were no serious side effects from aerosol amphotericin B prophylaxis, accrual in the study will continue for a total of 380 patients.