There is considerable evidence that susceptibility to idiosyncratic drug-induced liver injury (DILI) is genetically determined. Though genetic associations with DILI have been reported since the 1980s, the development of genome-wide association studies has enabled genetic risk factors for DILI, in common with other diseases, to be detected and confirmed more confidently. Human leukocyte antigen (HLA) genotype has been demonstrated to be a strong risk factor for development of DILI with a range of drugs and the underlying mechanism, probably involving presentation of a drug-peptide complex to T cells is increasingly well understood. However, specific HLA alleles are not associated with all forms of DILI and non-HLA genetic risk factors, especially those relating to drug disposition, also appear to contribute. For some drugs, there is evidence of a dual role for HLA and drug metabolism genes. Though the associations with non-HLA genes have been less well replicated than the HLA associations, there is increasing evidence that drug metabolism genes such as NAT2 and UGT2B7 contribute to some forms of DILI. Translating current genetic findings on DILI susceptibility to the clinic has been relatively slow, but some progress is now being made. In the future, DNA sequencing may lead to the identification of rare variants that contribute to DILI. Developments in the related area of epigenomics and in the development of improved models for DILI by use of genetically defined induced pluripotent stem cells should improve understanding of the biology of DILI and inform drug development.