OBJECTIVES: Poor medication adherence is common in diabetes potentially causing poor health outcomes and complications.  The aim of this study was to analyze the discontinuation rate of initiated basal insulin analog in type 1 and type 2 diabetic patients in Finland.  METHODS: The data was obtained from the national reimbursement registry. Study population consisted of 14 462 diabetic patients (18% had type 1 diabetes) who started basal insulin analogs (insulin glargine or insulin detemir) in 2012. Patients were followed by their insulin purchases for 18 months after the initiation. The data was analysed with χ-test and logistic regression analysis. Logistic regression analysis was used to find out what variables (age, gender, type of diabetes, type of insulin analog) explain patient staying in the treatment.  RESULTS: Type of insulin, gender, age and type of diabetes had statistically significant influence on patients’ treatment adherence (p < 0.001 for all). Overall 47 % of patients starting insulin detemir and 39 % starting insulin glargine patients discontinued their basal insulin treatment within 18 months of the initiation. Most of the patients stopped treatment within first 6 months after the initiation. In type 1 diabetic patients, 42 % of insulin glargine patients and 57 % of insulin detemir patients stopped the initiated treatment. In type 2 diabetic patients, 35 % stopped insulin glargine and 38 % insulin detemir. In only 15% of the patients discontinuing the initiated basal insulin, death or switch to other insulin or GLP-1RA explained the discontinuation suggesting non-adherence to insulin therapy from other reasons.  CONCLUSIONS: There is a considerable proportion of diabetic patients discontinuing their initiated basal insulin analog. Future studies are warranted to examine the detailed reasons for discontinuation.