In this study of 100,949 new users of oral bisphosphonates age >/= 35 years, "early quitters" were found to differ from others with poor refill compliance in terms of socioeconomic, demographic, and treatment-related characteristics. New risk factors for poor compliance and persistence were identified. INTRODUCTION: Poor compliance with anti-osteoporotic therapy is an on-going worldwide challenge. In this study, we hypothesized that "early quitters" differ in socioeconomics, demographics, co-medications, and comorbid conditions from other patients with low compliance. METHODS: The study was a register-based nationwide cohort study of anti-osteoporotic therapy comprising 100,949 men and women. Statistical analysis including backward stepwise logistic regression analysis was used to explain causes of treatment failure and Kaplan-Meier survival analysis to estimate persistence of treatment. RESULTS: It was noted that 56.6 % of the patients were persistent and compliant, 4.7 % of the patients were persistent but "low compliant" while 38.7 % of the patients were "early quitters". "Early quitters" were found to differ in socioeconomics from "low compliant" patients. Differences concerning increased risk of "early quitters" were associated with high household income, subjects' age 71.9-79 years, living in the countryside or village, prior treatment with analgesics and anti-parkinson drugs, and dementia. Differences concerning decreased risk of "early quitters" were associated with male, living in an apartment, children living at home, living close to a university hospital, anti-osteoporotic therapy other than alendronate, number of drugs especially above three, pulmonary disease, collagen disease. CONCLUSION: The results suggest a need for improved support for patients to facilitate the interpretation of the disease and the perception of the benefits and risks of treatment-to reduce the risk of "early quitters". We were able to identify new risk groups that may be candidates for targeted actions.