Central nervous system (CNS) relapse is a challenging complication in patients with diffuse large B-cell lymphoma (DLBCL). Thus, identification of the high-risk population, in whom prophylactic treatment may play a significant role, is critical. We calculated the incidence of CNS relapse and evaluated the risk factors for CNS relapse using competing risk regression analysis. A total of 386 patients who received CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone) with (n = 203) or without (n = 183) rituximab were analyzed. The 5-year cumulative incidence of CNS relapse was 6.7%. Multivariate analysis identified three independent risk factors: bulky disease (subhazard ratio [SHR] 3.34, 95% confidence interval [1.45-7.66], p = 0.004), absolute lymphocyte count <1.0 × 10(9)/L (SHR 2.38 [1.05-5.39], p = 0.037) and extranodal involvement (SHR 2.90 [1.01-8.33], p = 0.047). Patients with three risk factors represented 6% of patients, in whom the 5-year cumulative incidence was 26%. Larger scale studies are needed to validate our results. A better management strategy in patients with high-risk disease is critically needed.