The objective of this population-based case-control study was to identify risk factors for hip fracture among elderly women and men, particularly factors during young and middle adult life. The study base comprised people aged 65 years and over living in a defined region in Sydney, Australia, during 1990-1991. Cases were recruited from 12 hospitals, and controls were selected using an area probability sampling method, with additional sampling from nursing homes. There were 416 subjects (209 cases and 207 controls); proxy respondents were needed for 27 percent of the subjects. Smoking, underweight in old age, overweight at age 20 years, and weight loss were associated with an increased risk of hip fracture. Consumption of dairy products, particularly at age 20 years, was associated with an increased risk of hip fracture in old age. Multivariate-adjusted odds ratios for quintiles of dairy product consumption at age 20 years were 1.0 (lowest quintile), 0.8, 1.8, 3.4, 2.9 (highest quintile). Caffeine and alcohol intake were not associated with hip fracture risk. Some of the results of this study were unanticipated and may be due to chance or bias. If confirmed by other studies, these results would challenge some current approaches to hip fracture prevention.