We compared the incidence of bacterial pneumonia among 336 patients who discontinued trimethoprim-sulfamethoxazole (TMP-SMX) as prophylaxis against Pneumocystis carinii pneumonia (PCP) with that among 75 patients who fulfilled the criteria for discontinuation but continued receiving prophylaxis. The difference in the overall incidence rates for the 2 groups (1.2 events per 100 person-years) was not statistically significant. Discontinuation of TMP-SMX prophylaxis against PCP is not associated with a significant increase in the incidence of bacterial pneumonia among patients with a sustained CD4 cell count increase to >200 cells/microL.