There is convincing epidemiological evidence for a strong hereditary component to hand osteoarthritis (OA). For example: (1) greater concordance for hand OA in monozygotic than in dizygotic twins, the estimated proportion of variance explained by genetic factors being as high as 0.59; and (2) a substantially increased risk of hand OA in first-degree relatives (siblings, parents, offspring) of subjects with hand OA. Such evidence clearly justifies a search for the genes involved. However, gene association studies in genetically complex polygenic conditions such as OA present many problems, including case definition, late age of phenotype expression and adjustment for other constitutional and environmental risk factors. Nevertheless, association studies of affected sibling pairs and nuclear families, using candidate gene and genome wide screening and transmission disequilibrium testing, suggest no association with candidates such as COL2A1 (responsible for some rare monogenic syndromes of premature generalized OA) but possible associations, currently not isolated, on chromosome 2q. Such ongoing work and subsequent gene-gene and gene-environment interaction studies are likely to give important, perhaps unexpected, insights into the pathogenesis of hand OA.