Enzymes of phase I (cytochromes P450) and phase II (UDP [uridine diphosphate]-glucuronosyltransferases) of drug metabolism are targets of autoimmunity in the following chronic liver diseases of different etiology: 1)autoimmune hepatitis (AIH); 2) hepatitis associated with the autoimmune polyendocrine syndrome type 1 (APS-1); 3) virus-induced autoimmunity; and 4) drug-induced hepatitis. AIH is diagnosed by the following: the absence of infection with hepatitis viruses; the presence of a threshold of relevant factors, including circulating autoantibodies, hypergammaglobulinemia, female sex (female/male ratio 4:1), human leukocyte antigen (HLA) B8, DR3, or DR4; and benefit from immunosuppression. Patients with autoimmune hepatitis type 2 (AIH-2) are characterized by antibodies directed against liver and kidney microsomes, by an early onset of autoimmune hepatitis, which is a more aggressive course of the disease, and by a higher prevalence of autoimmunity directed against other organs. The major target of autoimmunity in patients with AIH-2 is cytochrome P450 2D6. Epitope mapping experiments revealed four short linear epitopes on cytochrome P450 2D6, recognized by liver/kidney microsomal autoantibodies type 1 (LKM-1) in patients with AIH-2. In addition, about 10% of the patient sera contain autoantibodies that detect a conformational epitope on UDP-glucuronosyltransferases (UGTs) of family 1. Presently, LKM-1 autoantibodies are used as diagnostic markers for AIH-2. It is unclear whether these autoantibodies have a pathogenetic role. Hepatitis is found in some patients with APS-1. Presumably this also is an autoimmune liver disease. APS-1 patients with hepatitis may develop autoantibodies directed against microsomal P450 enzymes of the liver; however, these autoantibodies do not recognize cytochrome P450 2D6, but they do recognize cytochrome P450 1A2. Autoimmunity in patients with APS-1 usually is directed against several organs simultaneously, and several organ specific autoantibodies may exist. Interestingly, APS-1 patients may produce various anti-cytochrome P450 antibodies. In addition to the hepatic anti-cytochrome P450, 1A2 autoantibodies are directed against steroidogenic cytochromes P450, namely P450 c21, P450 scc, and P450 c17. These autoantibodies correlate with adrenal and ovarian failure and often these steroidal cell autoantibodies precede the manifestation of adrenal or ovarian dysfunction. Whether anti-P450 1A2 autoantibodies have a similar predictive value is not yet known. LKM autoantibodies are further found in association with chronic hepatitis C and D. In chronic hepatitis C, the major target of LKM autoantibodies is cytochrome P450 2D6. Predominantly, conformational epitopes are recognized by LKM-1 sera of patients with chronic hepatitis C. In 13% of patients with chronic hepatitis D, LKM-3 autoantibody is detectable. The target proteins are UGTs of family 1 and in a minority of sera UGTs of family 2. The epitopes are conformational. All hepatic diseases discussed earlier have in common that autoimmunity, which is directed against enzymes of drug metabolizing multigene families. Each disease is characterized by a specific pattern of autoantibodies, with apparently little overlap. For example, LKM-1 autoantibodies, which are directed against P450 2D6, seem to overlap between AIH and chronic hepatitis C. However, a close examination of these autoantibodies shows differences between LKM-1 autoantibodies from patients with chronic hepatitis C and with AIH. In AIH, LKM autoantibodies are more homogenous, titers are higher, and major autoepitopes on cytochrome P450 2D6 are small and linear. LKM autoantibodies in viral hepatitis C are more heterogeneous and there are multiple epitopes, many of which are conformational. These differences indicate the different mechanisms that are involved in the generation of autoimmunity. Some forms of drug-induced hepatitis manifest as an attack of the immune system against hepatocytes; thus, it is important to distinguish toxic liver damage from such immunoallergic conditions, termed immune-mediated drug-induced hepatitis. Hepatotoxicity usually is the damage that a drug causes directly to the liver, e.g., via adduct formation or DNA injury; it manifests immediately and in a dose-dependent manner. Liver damage in immunemediated drug-induced hepatitis occurs after a significant lag-period; it is dose independent, does not occur after the first exposure, and is characterized by autoantibodies. In hepatitis caused by tienilic acid and dihydralazine, it is believed that these drugs are metabolically activated by cytochromes P450, resulting in adduct formation and the generation of a novel domain on the protein. This novel, modified self-domain may be recognized by T-cells and may induce an immune attack directed against cytochrome P450-bearing hepatic cells. Similar processes also seem to be involved in anti-convulsant induced hepatitis, and recent results indicate that autoimmune processes may even be involved in the development of alcoholic liver disease (ALD). In contrast, little is known about the mechanisms of autoantibody formation in autoimmune- and in virus-induced autoimmunity. The following questions must still be answered: (1) Which is the role of adduct formation?; (2) Is there a role of molecular mimicry between host and viral proteins in the induction of hepatic autoimmunity?; and (3) Can the high prevalence of nonhepatic autoimmune diseases in AIH-2 be explained by spreading of autoreactivity from the liver to other organs by cross-reactive epitopes? Learning more about the fine specificity of autoantibodies and their epitopes may improve diagnosis and management of patients and finally help us to understand mechanisms of autoimmunity.