The synovitis of rheumatoid arthritis (RA) was long regarded merely as an unspecific chronic inflammatory process of minor diagnostic value and therefore did not play a major role in the understanding of the pathogenesis of RA. It is only in recent years, along with the observation that T and B cells are expanded oligoclonally in synovial tissue and that B cells are able to undergo a local germinal center (GC) reaction, that the synovial tissue has come to be regarded as a site of specific immune processes. The analysis of the immunoglobulin (Ig) gene repertoire had great impact on the understanding of B cell response in lymphatic organs and was subsequently applied to B cells from RA patients. The analyses of the variable (V) regions of the Ig heavy (H) and light (lambda) chains suggested that an antigen specific activation and differentiation of B cells into plasma cells (Plc) takes place in the chronically inflamed synovial tissue of patients with RA. It seems that in a subset of RA patients the synovial tissue develops into an ectopic lymphoid tissue that supports a local GC reaction. Ectopic GC are characteristic of RA; however, they are in general absent from synovitis of osteoarthritis (OA). Here the accumulation of Plc follows a different mechanism. Highly mutated VH genes suggest that in OA memory B cells migrate into the synovial tissue with subsequent differentiation into Plc but without further V gene diversification. Therefore in synovitis two patterns of B cell activation can be differentiated: the maturative and the accumulative type. These two patterns are not definitely disease linked. The maturative type is only found in RA whereas the accumulative type occurs in both diseases. Clinically RA is defined via serum antibodies to the constant region of Ig, so-called rheumatoid factor. However, the spectrum of autoreactive B cells in RA patients is wide and is based on the study of antibody specificities in serum, in synovial fluid and B cell lines derived from peripheral blood, bone marrow, synovial fluid and synovial tissue. These analyses defined non-organ-specific and organ-specific antigens. One can reasonably assume that the disease is far too complex to be explained by only a single antigen. There is a whole combination of antigens acting in a multistep manner that is responsible for RA pathogenesis. It can be hypothesized that chronic synovitis, which is the underlying mechanism of joint destruction, follows a three-step process: (a) initiation, (b) destruction, and (c) perpetuation. The characterization of antigens driving the local synovial B cell maturation and accumulation could lead to an understanding of the process perpetuating the disease. Identification of arthritogenic antigens may yield new avenues for diagnostics and immunotherapy but also a new approach for prevention by vaccines with antigens probably defined by synovial B cell reactivity.