The casewise concordance for grade I & II OA changes in monozygotic (MZ) and dizygotic (DZ) twins was higher than would be expected from the prevalence in the group as a whole, suggesting familial clustering. Concordance for OA among MZ twins was more than DZ twins suggesting a genetic basis for hip OA. The same was for true for joint space narrowing. There was no excess of sclerosis (=5mm) in the MZ twins, although both MZ and DZ concordance rates were double the expected prevalence, suggesting familial aggregation without genetic basis. Amongst potential confounding factors, only age and body mass index were contributed to risk of hip OA. A greater environmental sharing between MZ twins compared with DZ twins was found; factors included smoking, alcohol consumption, and life-time exposure to exercise. None of these factors appeared to be significantly associated with OA.