OBJECTIVES: Management of Spanish patients with Type 2 diabetes and predictors of adherence to their medication. METHODS: Multicenter, naturalistic study of type 2 diabetes patients with a diagnosis longer than one year. Patients were consecutively included in the study among those attending any of the selected 30 Primary Care Centers distributed throughout Spain Sociodemographic and clinical variables were collected. Patients completed the Morisky-Green questionnaire as a measure of adherence. A logistic regression was performed with sociodemographic, clinical, and treatment variables to predict factors of adherence. RESULTS: A total of 294 patients were included in the study (mean age=67.5yrs; mean duration of diabetes=9.9yrs). A total of 6.1% were not receiving anti-diabetic pharmacological treatment, 70.4% received oral(s) therapy, 10.9% were in insulin therapy, and 12.6% a combination of insulin and oral(s) therapy. A total of 58.2% of patients had poor glycemic control (HbA1c>6.5%). With respect to diabetic complications, 20.4% of patients suffered from microvascular complications only, 11.2% reported macrovascular issues only, while 14.3% suffered from both micro and macrovascular complications. The Morisky-Green questionnaire showed only 47.8% of patients under pharmacological treatment reported a high level of adherence. The logistic regression models showed in that in those patients receiving only insulin, the probability of having a high level of adherence is significantly higher (odds ratio:2.7) than non-insulin receiving patients. Moreover, when a patient's HbA1c value is lower by one percentage point, the probability of having a high level of adherence is greater (odds ratio:1.3) compared to the adherence seen at a higher A1c level. CONCLUSIONS: Oral agents were the most commonly prescribed anti-diabetic pharmacological agent. Less than half of the patients reported high adherence to their medication. The level of diabetes control and insulin treatment were found to be associated with adherence although further work should investigate the causal relationship between these variables.