OBJECTIVES: Severe infections contribute to morbidity and mortality in antineutrophil cytoplasm antibody-associated vasculitis (AAV). This study aimed to identify risk factors associated with severe infections in participants of the Rituximab versus Cyclophosphamide for ANCA-Associated Vasculitis (RAVE) trial. METHODS: Data on 197 patients recruited into the RAVE trial were analysed. Participants received either rituximab (RTX) or cyclophosphamide (CYC), followed by azathioprine (AZA). Clinical and laboratory data of patients with and without severe infections (≥grade 3, according to the Common Terminology Criteria for Adverse Events version 3.0) were compared. Risk factors for severe infections were investigated using Cox-regression models. RESULTS: Eighteen of 22 (82%) severe infections occurred within 6 months after trial entry, most commonly respiratory tract infections (15/22, 68%). At baseline, lower absolute numbers of CD19+ cells were observed in patients with severe infections either receiving RTX or CYC/AZA at baseline, while CD5+B and CD3+T cells did not differ between groups. In Cox-regression analysis, higher baseline serum immunoglobulin M levels were associated with the risk of severe infections, whereby a higher baseline total CD19+B cell number and prophylaxis against Pneumocystis jirovecii with trimethoprim-sulfamethoxazole (TMP/SMX) with decreased risk of severe infections. Use of TMP/SMX was associated with lower risk of severe infections in both groups, receiving either RTX or CYC/AZA. CONCLUSIONS: The use of low-dose TMP/SMX is associated with reduced risk of severe infections in patients with AAV treated with either RTX or CYC/AZA. Reduced B cell subpopulations at start of treatment might be a useful correlate of reduced immunocompetence.