In a prospective study, 50 consecutive women with fractured hips were compared for body mass, skeletal size, handgrip strength and the 12-point information/orientation CAPE score with 50 age-matched women who had never broke their hips. Women were excluded from both groups if they lived in institutions or were severely disoriented. Although weight and skeletal size were significantly lower in the patients than in the controls, weights when corrected for skeletal size were not significantly different. Handgrip was significantly lower in the cases than in the controls. Both handgrip strength and CAPE score were significant correlates of hip fracture in multiple regression analysis. After exclusion of heavily dependent patients, hip fracture may be associated with reduced muscle strength rather than reduced body mass or fat.