Immune recognition of self-proteins features prominently in the early pathogenesis of autoimmune rheumatic diseases such as rheumatoid arthritis (RA), Sjögren's syndrome (SS), systemic lupus erythematosus (SLE) and systemic sclerosis. The mechanisms which provide lymphocytes with access to such autoantigens are therefore fundamental in creating the opportunity for autoimmune responses to develop. It has long been thought that the tissue or cellular location of some self-proteins may determine that they are normally 'hidden' from immune recognition, thereby reducing their potential for autoantigenicity. Recently, this concept has been extended to apply even to different epitopes within the same protein. Many studies, encompassing a wide variety of antigens, have shown that some epitopes are not presented for recognition by T lymphocytes unless they are produced in unusually large concentrations or unless they are freed from the configuration of their native antigen. Epitopes for which this phenomenon occurs are described as cryptic. There is increasing interest in the possibility that crypticity may be an important characteristic of epitopes which are recognized by T lymphocytes in autoimmune pathogenesis. The evidence which has led to this theory and its significance are reviewed.