Autoimmune hepatitis (AIH) is mediated by a T-cell attack upon liver parenchyma. Susceptibility to the development of AIH is genetically determined. While particular MHC haplotypes are known risk factors, it has been widely speculated that autoimmune liver damage can be regulated by additional genetic loci unlinked to MHC. However, evidence for the existence of such loci in humans is scant. We examined the contribution of the MHC in a murine model of autoimmune hepatocellular injury. BALB/c mice lacking the immunoregulatory cytokine transforming growth factor-beta1 (TGF-beta1) rapidly develop autoimmune T-helper 1-mediated necroinflammatory liver disease. Susceptibility to liver damage is strictly regulated by genetic background. Whereas TGF-beta1-deficient mice on the BALB/c background develop necroinflammatory liver disease, TGF-beta1-deficient mice on the 129/CF-1 genetic background do not. We asked whether MHC locus haplotype is the principal determinant of genetic susceptibility to liver disease in this model system. BALB/c mice harbor the H-2d haplotype. We used a 'haplotype swapping' approach to generate H-2b or H-2k congenic BALB-background TGF-beta1-deficient mice. In addition, F1 (BALB/c x 129/CF-1)-TGF-beta1-deficient mice were generated. As determined by plasma transaminase levels and histopathology, severe necroinflammatory liver disease developed in all BALB-background TGF-beta1-deficient mice, regardless of H-2 haplotype, but developed neither in 129/CF-1-TGF-beta1-deficient mice nor in F1 (BALB/c x 129/CF-1)-TGF-beta1-deficient mice. Thus, H-2d is neither necessary nor sufficient for the development of necroinflammatory liver disease in BALB-background TGF-beta1-deficient mice. This constitutes the first direct evidence that susceptibility to autoimmune hepatocellular damage, at least in mice, can be determined by genetic loci distinct from the MHC.