This study evaluated reasons for healthcare expenditures both before and after the occurrence of fractures among Medicare beneficiaries. In a previous study we examined healthcare expenditures in the 6 months before and after fractures. The difference-"incremental" expenditures-provides one estimate of the potentially avoidable costs associated with fractures. We constructed a second estimate of the cost burden-"attributable" expenditures-using only those costs recorded in claims with fracture diagnosis codes. Attributable expenditures accounted for only 24% to 60% of incremental expenditures, depending on the fracture site. We examined health care expenditures between 1999 and 2005 among Medicare beneficiaries who experienced fractures (cases) and among beneficiaries who did not experience fractures (controls), matched to cases on age, race, and sex. We also examined healthcare expenditures for cases and controls for 24 months prior to the fracture index date. When expenditures associated with diagnoses for aftercare, joint pain, and osteoporosis, other musculoskeletal diagnoses, pneumonia, and pressure ulcers were included, the proportion of incremental costs directly attributable to fracture care rose to 72% to 88%. Expenditures prior to fracture were higher for cases than controls, and the rate of increase accelerated over the 12 months prior to the hip fracture. Our findings confirm that the original incremental cost analysis constituted a satisfactory method for estimating avoidable costs associated with fractures. We also conclude that those with fractures had much higher and growing healthcare expenditures in the 12 months prior to the event, compared with age-, race-, and sex-matched controls. This suggests that patterns of healthcare services utilization may provide a means to improve fracture prediction rules.