This case-control study examined the effect of diet on the risk of postmenopausal fracture of the hip and wrist. Cases, women aged 50-84 years, were admitted to one of four Metropolitan Toronto hospitals during the period September 1983 through May 1985. Controls were women of the same age, admitted to the same hospitals, and seen for orthopaedic or general surgical complaints. Data were collected through the use of a standardized structured questionnaire administered by trained female interviewers, and included dietary, sociodemographic, medical, and behavioural variables. Data analysis was conducted on 102 hip fracture cases, 154 wrist fracture cases, and 277 controls. Higher dietary calcium intake only slightly increased the risk of hip fracture; however, it was associated with a significantly decreased risk for fracture of the wrist, at the level > or = 1 g/day. Coffee and tea consumption appeared to be unrelated to fracture risk. Finally, alcohol intake was associated with a statistically nonsignificant increased risk in both fracture types. These findings point to the importance of dietary factors on fracture risk, and indicate a need for future studies to stratify analysis on the basis of fracture type.