Low persistence with osteoporosis medication is associated with higher fracture risk. Previous studies estimated that 1-year persistence with osteoporosis medication is low. Our aim was to study persistence with osteoporosis medication among patients with long-term follow-up (to 5 years). The InterAction Database (IADB) was used to analyze persistence of 8610 Dutch patients initiating osteoporosis drugs between 2003 and 2011. Drugs under study were alendronate, risedronate, ibandronate, etidronate, raloxifene and strontium ranelate. Cumulative persistence rates were calculated after different time frames (3 months-5 years) using survival analysis. Multivariate Cox proportional hazard analyses were used to identify determinants of non-persistence. Furthermore, switching rates of persistent patients who initiated bisphosphonate therapy were analyzed. Persistence with osteoporosis therapy was 70.7 % (95 % CI, 69.7-71.7), 58.5 % (95 % CI, 57.4-59.6 %), 25.3 % (95 % CI, 24.1-26.5) after 6 months, 1 and 5 years, respectively. Determinants associated with higher risk to non-persistence within the first year were daily dosing regimen [HR, 1.76 (95 % CI, 1.46-2.14)], age <60 years [HR, 1.26 (95 % CI, 1.19-1.34)] and use of glucocorticoids [HR, 1.16 (95 % CI, 1.07-1.26)]. Monthly dosing schedule and use of generic brands of alendronate did not show a significant association with non-persistence. Approximately 4.0 % of patients initiating therapy with weekly alendronate or weekly risedronate switched therapy. Persistence with osteoporosis medication is low. Because low persistence is strongly associated with higher fracture risk, interventions to improve persistence are recommended. This study identified several patient groups in whom such interventions may be most relevant.