A variety of clinical, epidemiological, and experimental data suggest that rheumatic heart disease and autoimmune myocarditis are not only similar in their pathogenesis, but may often be due to combined infections with coxsackie virus (CX) and streptococcus A bacteria (SA). This paper reviews the evidence for this hypothesis, provides some new experimental data supporting the hypothesis, and suggests specific experiments for testing it. While, it is well-established that the M protein of SA mimics myosin, we demonstrate using homology search tools that various CX proteins mimic actin. We further demonstrate that antibody against CX recognizes actin as an antigen, and that anti-actin antibodies recognize CX antigen. Thus, anti-CX antibodies may also target muscle. Moreover, since myosin and actin are molecularly complementary, it follows that some SA and CX proteins may be molecularly complementary. Some antibodies against these complementary proteins in SA and CX should therefore act like idiotype-antiidiotype antibodies. We show that, indeed, CX and SA antibodies precipitate each other. Thus, it is possible that combined CX-SA infections produce more severe disease by producing pairs of idiotypic antibodies that act like antiidiotypic antibodies as well, thereby, disregulating immune control and triggering an autoimmune reaction against both myosin and actin simultaneously. We predict that combinations of the appropriate actin- and myosin-like antigens from CX and SA will, therefore, be much more autoimmunogenic than antigens from CX or SA alone, and that the combination will not require use of adjuvants or self-proteins that many current protocols require. It is possible that co-infections involving CX or SA with other infectious agents may produce similarly enhanced disease.