Trimethoprim-sulfamethoxazole (TMP/SMX) is used in children with acute lymphoblastic leukemia (ALL) to prevent Pneumocystis pneumonia (PCP). We explored to which extent TMP/SMX influenced methotrexate (MTX)/6-mercaptopurine (6MP) dosage, myelosuppression, and event-free survival (EFS) during maintenance therapy. Of 447 study patients treated by the NOPHO ALL92 protocol, 120 patients received TMP/SMX continuously for 2–7 d/wk (TMP/SMX2–7) and 287 patients never received TMP/SMX (TMP/SMXnever). Ten patients (all TMP/SMXnever) developed PCP, eight of which occurred within 7 months from the start of maintenance therapy. The TMP/SMX2–7 group received lower oral 6MP doses than TMP/SMXnever patients (50.6 vs. 63.9 mg/m2/d; P < 0.001) but had lower absolute neutrophil counts (ANC) (median 1.7 vs. 2.0 × 109/L; P < 0.001). In Cox multivariate analysis, higher ANC levels (P = 0.04) and male gender (P = 0.06) were related to reduced EFS. ANC had no effect on EFS among TMP/SMX2–7 patients (P = 0.40) but did for TMP/SMXnever patients (P = 0.02). The difference in the effect on EFS between TMP/SMX2–7 and TMP/SMXnever patients was not significant (P = 0.46). EFS did not differ between TMP/SMX2–7 and TMP/SMXnever patients (0.83 vs. 0.83; P = 0.82). These results suggest that TMP/SMX is effective in preventing PCP and may have an antileukemic effect. TMP/SMX should be given the entire duration of maintenance therapy