In rheumatoid arthritis (RA) there are currently no good indicators to predict a clinical response to rituximab. The purpose of this study was to monitor and determine the role of peripheral blood cytokine profiling in differentiating between a good versus poor response to rituximab in RA. Blood samples were collected at baseline and at 3 months from 46 RA patients who were treated with rituximab. Responders are defined by the presence of three of four American College of Rheumatology criteria: >or=20% decrease in C-reactive protein, visual analogical score of disease activity, erythrocyte sedimentation rate and improvement of the disease activity score (28) (four values) by >or=1.2 obtained at 3 months. Twelve cytokines were measured from serum collected on days 0 and 90 by proteomic array, including interleukin-6 (IL-6), tumour necrosis factor-alpha, IL-1a, IL-1b, IL-2, IL-8, interferon-gamma, IL-4, IL-10, monocyte chemoattractant protein-1, epidermal growth factor and vascular growth factor. We showed that C-reactive protein and IL-6 levels decrease significantly at 3 months in the responder group compared with baseline. At day 90 we identified a cytokine profile which differentiates responders and non-responders. High serum levels of two proinflammatory cytokines, monocyte chemoattractant protein-1 and epidermal growth factor, were significantly higher in the responder group at day 90 compared with non-responders. However, we were not able to identify a baseline cytokine profile predictive of a good response at 3 months. These findings suggest that cytokine profiling by proteomic analysis may be a promising tool for monitoring rituximab and may help in the future to identify responder RA patients.