Patients receiving maintenance dialysis are at a 5-fold risk of fracture compared to those with normal renal function, with attendant poor outcomes. Current risk prediction methods are limited to the application of scores used in the general population, which lack validation and produce risk prediction timelines that are not applicable to those receiving dialysis. The lack of a well validated, easy-to-use tool to identify individuals at high risk of fracture hinders both the clinical prevention of fractures as well as the inclusion of these individuals in therapeutic trials. I will present a risk prediction model derived from a population-based cohort of over 11,000 adults receiving maintenance dialysis and 839 fractures observed in Ontario, Canada. This model uses commonly available variables including age, sex, previous fracture, current steroid use, proton pump inhibitor use, calcitriol use, length of time receiving dialysis, and the concentrations of PTH and serum albumin. Discrimination values were good (c-statistic of 0.72 at 3 years). I will then discuss the future implications of a risk model including its inclusion in future studies of interventional fracture risk prevention studies