This review will mainly highlight data from selected, independent studies which collectively implicate a primary role for T cells in the pathogenesis of collagen arthritis in rats. Conferring insusceptibility to this experimental disease with the use of polyclonal, T cell specific antiserum provided direct initial evidence for this conclusion. Substantiation for the theory of a dominant T cell role in collagen arthritis was afforded by T cell line vaccination; scrutiny showed that the mechanism accounting for this protection was a specific down-regulation of the cellular response to collagen. Additional support came from experiments which showed that as few as 103type II collagen specific T line cells were capable of provoking a sustained proliferative synovitis when instilled into the knee joint cavity of syngeneic naive rats. Further analysis of this phenomenon revealed that the arthritogenic capacity of various collagen-reactive line cells correlated with their ability to release a 65-Kd, collagen-binding lymphokine. This antigen-specific lymphokine was designated arthritogenic factor, based on an arthritogenic activity in the knee joint bioassay similar to that of the cells. A functional and physicochemically identical rat arthritogenic factor has also been identified in the adjuvant model of arthritis. These data support the premise that a major effector mechanism in experimental rat arthritis is the release of arthritogenic factor by expanded clones of autoreactive T cells; they also indicate that substantive efforts should be undertaken to seek to identify arthritogenic factor-like lymphokines in patients with chronic inflammatory synovial disease. As an equally plausible alternative hypothesis, the review will close with a brief discussion of recent findings supporting the possible involvement of cartilage-binding, complement-fixing anti-type II collagen antibodies in the pathogenesis of rheumatoid arthritis.