The development of highly effective biological therapies directed against T cells in several animal models of autoimmune disease has prompted trials of similar approaches in rheumatoid arthritis (RA). However, it is unlikely that these approaches will abrogate long-standing disease. Indeed, considerable evidence indicates that although T cells likely play a critical role in induction of RA, non-T-cell-dependent pathways become increasingly dominant as the disease progresses. According to this model, specific T-cell therapies are likely to be most effective in early disease, whereas individualized combinations of biologies targeted against pathways dominating in the recipient's synovium are more likely to be efficacious in established disease. © 1994.