OBJECTIVE To investigate whether the APOE 4 allele was associated with increased risk of hip fracture in an older community-based sample and whether such an increased risk was independent of dementia and history of falling. METHODS Case-control study nested within a prospective community study. METHODS The Monongahela Valley Independent Elders Survey (MoVIES), an ongoing prospective community study of older adults in southwestern Pennsylvania. METHODS A total of 899 MoVIES participants (63.9% women; mean age, 76.2 years, SD = 4.9 years), who provided both information on hip fractures and blood samples for genotyping. METHODS Interview questions regarding hip fractures and falls, polymerase chain reaction to determine APOE genotype, and clinical assessment using a standardized protocol to determine the presence or absence of dementia. RESULTS Twenty-five subjects reported having hip fractures in the year preceding screening interviews. Subjects with one or two APOE 4 alleles were twice as likely as subjects without an APOE 4 allele to report hip fractures (age-adjusted OR = 2.1, 95% CI: 0.9-4.7). Based on multivariate analysis, subjects with a history of falling were more likely to report hip fractures (OR = 4.7, 95% CI: 2.1-10.8). After adjusting for history of falls and diagnosis of dementia, subjects with an APOE 4 allele were still twice as likely to report hip fractures (adjusted OR = 2.1, 95% CI: 0.9 - 4.7). CONCLUSIONS The APOE 4 allele appears to be a risk factor for hip fracture, independent of the effect of dementia and falling. Theoretically, this may be mediated by alterations in vitamin K metabolism. Caution should be used in interpreting these results, because the 95% confidence intervals for the odds ratios include 1.