OBJECTIVES: Clinical trials have demonstrated that drug therapy can reduce osteoporosis-related fracture risk in women over 50 years by up to 40% provided they consistently take their medication for a year or more. Non-adherence with drug therapies not only limits the drugs' effectiveness, but could also be associated with a higher fracture risk. The objective of this study was therefore to estimate fracture risk in relation to adherence with osteoporosis medication in actual practice. METHODS: Demographic, prescription drug use, physician services and hospitalization information for females with osteoporosis who were dispensed an osteoporosis medication between 1996 and 2001 (entry date defined as first dispensing in this period) was obtained from the Saskatchewan Health data files. Adherence to treatment was defined as drug available to cover 80% of the time. Subsequent fractures were identified via hospitalizations or physician contacts with a relevant diagnostic or procedure code. The risk of fractures in relation to adherence was examined using a Cox proportional hazards model with time-dependent covariates. The impact of other patient characteristics, including age, having suffered a prior fracture, prior use of osteoporosis medication and steroids, was also examined. RESULTS: 11,249 women suffering from osteoporosis were identified with a mean age at the time of the index prescription of 68.4 years and average follow-up of 2.3 years with a fracture rate of 4.5% per year. Patients who adhered experienced a 16% lower fracture rate. The effect of adherence was maintained after controlling for other patient characteristics that independently predict the fracture rate, including interactions. CONCLUSIONS: These results indicate that improving adherence in actual practice will significantly decrease the osteoporosis-related fracture risk.