The efficacy and safety of pentamidine aerosol versus intravenous cotrimoxazole (trimethoprim-sulfamethoxazole) [TMP/SMX] in the treatment of slight to moderatePneumocystis cariniipneumonia in HIV-infected patients were investigated in a comparative study. The study was designed as an open, randomised, multicentre, prospective study for a collective of 60 patients. Pentamidine aerosol (600 mg/day) was tested against intravenous cotrimoxazole (TMP 20mg/SMX 100mg per kg). Major inclusion criteria were: arterial partial oxygen pressure (pO2) >60mm Hg and forced vital capacity (FVC1) >2L or >50%. 24-hour urine and serum samples were used to evaluate the pharmacokinetics of the tested medications in selected patients. A statistical evaluation was performed according to the ‘intent to treat’ principle after a 4-week follow-up.Case reports of 22 patients treated with pentamidine and 24 treated with TMP/SMX were evaluated. Because of a high frequency of side effects in the TMP/SMX group, the study was discontinued after 46 patients had completed the treatment cycles. However, the response rate in the 2 groups was virtually identical. Complete remission occurred in 16 pentamidine patients and in 10 TMP/SMX patients who received the full course of treatment. In the pentamidine group, 3 patients showed slight side effects; no severe side effects occurred. In the TMP/SMX group, slight side effects were observed in 9 patients, and severe side effects in 7 patients. The difference in the frequency of side effects, the main target variable, was statistically significant (p = 0.0003; 2-tailed, Fisher's exact test; power of comparison = 95%). Treatment success in the absence of side effects was observed in 15 pentamidine patients and in 3 TMP/SMX patients.The study suggests that pentamidine aerosol has advantages over standard intravenous TMP/SMX treatment in slight to moderatePneumocystis cariniipneumonia.