To assess the clinical benefit of rituximab for HIV-associated Castleman disease, 5 patients infected with HIV with histologic-proven Castleman disease were prospectively enrolled to receive 4 infusions of rituximab. Clinical and biologic parameters (C-reactive protein, CD19 cell count, Kaposi sarcoma-associated herpesvirus [KSHV] viral load in peripheral blood mononuclear cells) were assessed before and at different time points following rituximab infusions. Two patients died very quickly after the beginning of rituximab therapy with no effect on both KSHV viral load and CD19 cell count. Three of 5 patients were considered in complete remission with no more clinical symptoms related to Castleman disease with a follow-up of 4 to 14 months. In 2 cases, clinical remission correlated with a dramatic decrease of KSHV viral load and C-reactive protein levels and a transitory but sharp decrease of CD19 cell count. In 2 responders, we observed an aggravation of Kaposi sarcoma. Our preliminary results suggest that rituximab may be effective in controlling Castleman disease in a subset of patients, although it may exacerbate concomitant Kaposi sarcoma.