Alcoholics often sustain hip and other fractures. However, a detailed examination of the association between alcohol consumption and hip fractures has not been undertaken. Specifically, the effects of moderate alcohol intake, of alcohol consumption in the elderly, and of changes in consumption have not been studied. Using a retrospective cohort design, the authors evaluated this association in the population-based Framingham Heart Study cohort, a group studied over 35 years and queried repeatedly about their alcohol consumption. In 117,224 person-years of observation, 217 hip fractures occurred. Heavy current alcohol consumption (defined as seven or more ounces (207 ml or more) per week) was associated with a modestly increased risk of hip fracture for women (relative risk (RR) = 1.54) and for men (RR = 1.26) after adjustment for age. In a logistic regression analysis controlling for age, sex, weight, and smoking, current alcohol consumption was associated with a significant (p = 0.01) increase in risk of fracture. The relative risks at different ages were not uniform. For those aged less than 65 years, moderate (2-6 ounces (59.14-177.4 ml) per week) and heavy (seven or more ounces (207 ml or more) per week) were associated with a substantial and significant increased risk, but there was only a marginal and nonsignificant increase in risk in those aged 65 years or more. Examination of the effect of changing alcohol consumption over time indicated that among present heavy alcohol users, past light alcohol consumption reduced the risk (p = 0.03) of fracture, whereas in those with present light consumption, past heavy intake had little effect on fracture risk (p = nonsignificant). In conclusion, alcohol consumption, especially if long-term and heavy, increases the risk of hip fracture.