The addition of rituximab to cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) chemotherapy results in significant improvement in clinical outcome for individuals with non-HIV-associated aggressive B-cell lymphoma. To assess the potential risks and benefits of the addition of rituximab to CHOP for HIV-associated non-Hodgkin lymphoma (HIV-NHL) 150 patients receiving CHOP for HIV-NHL were randomized (2:1) to receive 375 mg/m(2) rituximab with each chemotherapy cycle (n = 99) or no immunotherapy (n = 50) in a multicenter phase 3 trial. The complete response rate (CR + CRu) was 57.6% for R-CHOP and 47% for CHOP (P = .147). With a median follow-up of 137 weeks, time to progression, progression-free survival, and overall survival times were 125, 45, and 139 weeks, respectively, for R-CHOP and 85, 38, and 110 weeks, respectively, for CHOP (P = not significant, all comparisons). Treatment-related infectious deaths occurred in 14% of patients receiving R-CHOP compared with 2% in the chemotherapy-alone group (P = .035). Of these deaths, 60% occurred in patients with CD4 counts less than 50/mm(3). Progression-free survival was significantly influenced by CD4(+) count (P < .001) and International Prognostic Index score (P = .022), but not bcl-2 status. The addition of rituximab to CHOP in patients with HIV-NHL may be associated with improved tumor responses. However, these benefits may be offset by an increase in infectious deaths, particularly in those individuals with CD4(+) lymphocyte counts less than 50/mm(3).