The effect of change in weight in later years on risk of osteoporotic fractures in uncertain. Prior studies have assessed the relationship between weight change and risk of hip fracture only and have not examined whether the association between weight loss and increased risk of hip fracture is because of voluntary or involuntary weight loss. To determine the association between weight change in later years and the risk of fractures in elderly women and to assess the effect of weight loss intention on this relationship. The association between weight change and fracture risk was assessed in 6754 ambulatory, non-black women aged 65 years or older enrolled in the Study of Osteoporotic Fractures who had measurements of weight performed at both the baseline and fourth examinations (mean, 5.7 years between examinations). These 6754 women were followed up for all incident nonspine fractures and frailty fractures (defined as fractures of the proximal femur, pelvis, and proximal humerus) occurring after the fourth examination (average follow-up, 19.5 months). All incident fractures were confirmed by radiographic report. During an average of 19.5 months after the fourth examination, 264 women (4%) had at least 1 nonspine fracture, including 83 women who suffered frailty fractures. After adjustment for age, women who lost weight between the baseline and fourth examinations had an increased risk of subsequent nonspine fracture (relative risk [RR] per 10% decrease in weight, 1.32; 95% confidence interval [CI], 1.11-1.55). This observed increase in the risk of nonspine fracture was entirely caused by an increase in the risk of frailty fracture (age-adjusted RR per 10% decrease in weight, 1.86; 95% CI, 1.42-2.43). Weight change was not significantly related to other types of nonspine fractures (age-adjusted RR per 10% decrease in weight, 1.09; 95% CI, 0.90-1.33). Further adjustment for cigarette smoking, physical activity, estrogen use, medical conditions, health status, body weight, femoral neck bone mass, and rate of change in calcaneal bone mass did not substantially alter the association between weight change and frailty fracture (multivariate RR per 10% decrease in weight 1.68; 95% CI, 1.17-2.41). Weight change was an even stronger predictor of risk of frailty fracture in those women who were not trying to lose weight (multivariate RR per 10% decrease in weight, 1.81; 95% CI, 1.26-2.61). Involuntary weight loss in later years substantially increases the risk of frailty fracture including hip fracture in elderly women. Weight change is not related to risk of other nonspine fractures in older women.