Background Glucocorticoids are widely used for the treatment of various rheumatic diseases and are one of the major risk factors for bone fragility associated fractures. However, beyond randomized clinical trials, there is little evidence in actual clinical practice. Objectives Our aim is to evaluate the appearance of osteoporosis and fractures in patients undergoing corticosteroid therapy. Methods We performed an observational retrospective descriptive study with Vasculitis (several subtypes) or Polymyalgia Rheumatica (PMR) patients diagnosed, who had received corticosteroid regimens with doses of at least 5 mg/day of prednisone (or equivalent) for a period of not less than 3 months, and in which the appearance of fractures was monitored for the next 2 years. Results We included 234 patients (62.4% females) with a mean age at the onset of corticosteroids of 75.8 years and mean duration corticosteroid regimens of 58 months. The most represented disease was PMR, with 183 patients (78.2%), followed by Large Vessel Vasculitis (12.4%), ANCA positive Small Vessel Vasculitis (4.3%), ANCA negative Small Vessel Vasculitis (3.4%), with a lower proportion of Medium Vessel vasculitis and other diseases such as Relapsing Polychondritis. Furthermore, 29 patients (12.39%) had previous fragility fractures, while 32 patients (13.68%) suffered one or more fractures after the start of corticosteroid therapy, being hip its most frequent location (4.7%), followed by vertebrae (3.8%). Likewise, the mean duration of corticosteroid treatment until the onset of new fractures was 4.5 months. Even without reaching statistical significance, we observed a greater age (80.5 years), a higher presence of the female gender (81.25%) and a previous fracture (18.75%) in patients with new fractures. In addition, 32.91% of patients had densitometric osteoporosis at the beginning of corticosteroid therapy. Comparison between different therapeutic options We observed an association between fractures and zoledronate treatment (P=0.012), and a tendency for ibandronate (P=0.07). The discreet entity of these differences could be explained by the statistical effect produced by the small number of new fractures, taking into account, moreover, the fact that in our Rheumatology Department we systematically carry out a fracture risk assessment to decide whether and when to start the treatment, and also prescribing more powerful therapeutic options to patients at higher risk of fracture. Finally, we did not observe significant differences in fracture-free time among the most frequent therapeutic options. Conclusion We observed a tendency to have a higher frequency of fractures in patients with previous fractures, older age and female gender. Less quantitative prevention of fractures was observed in patients with zoledronate and ibandronate, with no clear differences in fracture-free time. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests None Declared.