Purpose/introduction A Fracture Liaison Service (FLS) was set up at Lille University Hospital in 2016. The purpose of this study was to assess persistence with osteoporosis treatment in patients from the FLS over a period of 1 year, and to determine predictors of discontinuation. Methods The study population comprised adults of both genders, aged 50 or over, admitted to Lille University Hospital between January 2016 and January 2019 for a low-trauma fracture and managed in our FLS. Outcomes included (1) persistence rate at 1 year after treatment initiation, (2) persistence rate at 2 years after treatment initiation, (3) persistence rate at 1 and 2 years after treatment initiation according to type of treatment, (4) predictors of non-persistence, and (5) reasons for discontinuing treatment over 1 year after initiation. Persistence was determined using the Kaplan–Meier method. Results In all, 1224 patients (≥50 years old) with a recent history of low-trauma fracture (≤12 months) were identified. Of these, 380 patients – 79.2% female; mean (SD) age 76 (11) years – were seen at the FLS. In those 380 patients, 410 fractures were found and 360 of them (87.8%) were major fractures, breaking down as follows: vertebra (44%), hip (19%), proximal humerus (10%), and pelvis (8%). Osteoporosis treatment was prescribed for 367 (96.6%) patients and 275 of them began the prescribed treatment. The following anti-osteoporosis drugs were prescribed: zoledronic acid (n=150, 54.5%), teriparatide (n=63, 22.9%), and denosumab (n=39, 14.2%). Oral bisphosphonates were prescribed for a few patients (n=23, 8.4%). Persistence with osteoporosis medication (any class) was estimated at 84.1% (95% CI: 79.1% to 88.1%) at 12-month follow-up, and dropped to 70.3% (95% CI: 63.7% to 75.9%) at 24 months. When drug-specific analyses were performed using the Kaplan–Meier method, persistence rates at 12 and 24 months were found to be higher with denosumab than with any other treatment. Independent predictors of non-persistence at 12 months were ‘follow-up performed by a general practitioner (GP)’ – Odds Ratio (OR) for GP vs. FLS = 3.68; 95% CI, 1.52 to 8.90, p=0.004 – and ‘treatment with zoledronic acid’ – OR for zoledronic acid vs. denosumab = 3.39; 95% CI, 1.21 to 9.50, p=0.019; OR for zoledronic acid vs. teriparatide = 8.86; 95% CI, 1.15 to 68.10, p=0.035. Conclusions This study provides evidence of the success of our FLS in terms of long-term persistence with osteoporosis treatments. However, osteoporosis treatment initiation still needs to be improved.