Aetiology and pathogenesis of idiopathic dilated cardiomyopathy (DCM) are uncertain. The two major pathogenetic hypotheses are: 1) autoimmunity; 2) persistent viral infection. Indirect evidence for virus association comes from the finding of raised titres of antibody to coxsackievirus in DCM, but infectious virus has never been isolated in myocardium from DCM patients. Bowles et al. using the slot-blotting technique reported that enteroviral RNA was commonly detectable in the myocardium of patients with myocarditis (53%) and with DCM (52%). Other groups using this as well as more refined hybridization techniques have failed to confirm such a high prevalence. Detection of enteroviral genomic RNA in cardiac tissue does not, however, imply active infection or pathogenicity. Thus the mechanisms of chronic myocardial damage in the absence of whole competent infectious virus remain uncertain. The other major pathogenetic hypothesis in DCM involves autoimmune mediated damage to myocytes. Circulating organ specific autoantibodies have been reported in a quarter of a group of patients with idiopathic DCM. This suggests that there may be autoimmune mechanisms operating at least in this subset of patients, but the exact relation of these antibodies to the pathogenesis and prognosis needs to be defined. The abnormal expression of major histocompatibility complex class II antigens on cardiac microvascular endothelium in endomyocardial biopsy tissue from DCM patients, and the reported association with HLA-DR4 phenotype lend further support to the autoimmune hypothesis. The viral and the autoimmune hypothesis in chronic myocarditis and in DCM are not mutually exclusive. In experimentally murine virus-induced myocarditis infectious virus can no longer be recovered from the myocardium after two weeks, although nucleic acid sequences of the viral genome are still detectable. The development of chronic inflammation takes place only in mice with a predisposing genetic background. Chronic myocyte damage is associated with the production of circulating heart-specific autoantibodies and autoreactive lymphocytes. In this animal model chronic myocarditis appears to be a virus-triggered or precipitated autoimmune disease, rather than a persistent viral infection with tissue damage due to active virus synthesis and replication. A similar transition from acute myocarditis into DCM may occur in man.