Three new findings are reviewed that help to understand the mechanisms of action of anti-rheumatic gold drugs, such as disodium aurothiomalate (Na2Au(I)TM): i) We found that Na2Au(I)TM selectively inhibits T-cell receptor-mediated antigen recognition by murine CD4+ T-cell hybridomas specific for antigenic peptides containing at least two cysteine residues. Presumably, Au(I) acts as a chelating agent forming linear complexes (Cys-Au(I)-Cys) which prevents correct antigen-processing and/or peptide recognition by the T-cell receptor, ii) We were able to show that Au(I) is oxidized to Au(III) in mononuclear phagocytes, such as macrophages. Because Au(III) rapidly oxidizes protein and itself is re-reduced to Au(I), this may introduce an Au(I)/Au(III) redox system into phagocytes which scavenges reactive oxygen species, such as hypochlorous acid (HOCl) and inactivates lysosomal enzymes, iii) Pretreatment with Au(III) of a model protein antigen, bovine ribonuclease A (RNase A), induced novel antigenic determinants recognized by CD4+ T lymphocytes. Analysis of the fine specificity of these "Au(III)-specific" T-cells revealed that they react to RNase peptides that are not presented to T-cells when the native protein, i.e., not treated with Au(III), is used as antigen. The T-cell recognition of these cryptic peptides did not require the presence of gold. This finding has important implications for understanding the pathogenesis of allergic and autoimmune responses induced by gold drugs. Taken together, our findings indicate that Au(I) and Au(III) each exert specific effects on several distinct functions of macrophages and the activation of T-cells. These effects may explain both the desired anti-inflammatory and the adverse effects of antirheumatic gold drugs.