© 2018, Springer Science+Business Media, LLC, part of Springer Nature. The phase 3 teriparatide Fracture Prevention Trial showed significant reductions in vertebral (VF) and nonvertebral (NVF) fractures; however, patient exposure was insufficient for full analysis of low-incidence fractures, including hip. We assessed fracture results in pooled data from four prospective, observational teriparatide studies. Ambulatory women and men with osteoporosis received subcutaneous teriparatide 20 µg/day for up to 24 months per routine clinical practice. Fracture rates were compared between 6-month periods, using 0 to 6 months of treatment as the reference period. Analyses used a piecewise exponential model for first fracture. Hip, NVF, clinical VF (CVF), any clinical, and wrist fractures were assessed. For 8828 patients analyzed, mean age was 71 years; mean (SD) treatment duration was 17.4 (8.6) months. The rate of hip fracture decreased significantly for the > 12 to 18-month (− 47.7%) and > 18-month periods (-85.2%) versus the first 6 months of therapy, and for the > 18 versus the > 6 to 12-month period. NVF, CVF, and all clinical fractures were all significantly decreased in each post-reference period, with maximum decreases (> 18-month period) of 52.7%, 69.4%, and 61.2%, respectively, versus 0 to 6 months. No significant reduction was seen for rates of wrist fracture. Teriparatide treatment was associated with statistically significant decreases in hip fracture rate, particularly for > 18 months of treatment, and in NVF, CVF, and all clinical fracture rate in real-world patients. These results should be interpreted in the context of the non-controlled design of the source studies.