This study assessed the effect of aerosolized pentamidine prophylaxis on the clinical presentation and diagnostic sensitivity of induced sputum examination for Pneumocystis carinii pneumonia. Between January 1, 1988 and October 27, 1990, 348 induced sputum examinations were performed as the initial diagnostic procedure for P. carinii pneumonia in patients infected with the human immunodeficiency virus (HIV). Medical records were reviewed for all induced sputum examinations, and the study group consisted of patients who either had not received prophylactic therapy (n = 193) or had received aerosolized pentamidine prophylaxis (n = 126). A total of 29 induced sputum examinations in patients receiving either other prophylactic regimens or ongoing therapy for previously documented P. carinii pneumonia were excluded from the study group. A total of 72 consecutive episodes of P. carinii pneumonia were subsequently documented by induced sputum examination (n = 54), bronchoalveolar lavage (n = 16), thoracocentesis (n = 1), or autopsy (n = 1). A total of 44 episodes occurred in patients who had not received antipneumocystis prophylaxis, and 28 episodes occurred in patients who had received aerosolized pentamidine. Of patients capable of producing a sputum specimen for analysis, induced sputum examination had a significantly lower diagnostic yield of 64.3% in patients who had received aerosolized pentamidine prophylaxis compared with 92.3% in patients who did not receive prophylaxis (p less than 0.02, Fisher's exact test). When the data were analyzed on an intention to treat basis, although there was a trend suggesting a lower overall yield in the aerosolized pentamidine patients, the difference was not statistically significant (64.3 versus 81.8%, p = 0.17, Fisher's exact test).(ABSTRACT TRUNCATED AT 250 WORDS)