OBJECTIVES: To assess the treatment compliance and illness knowledge of diabetic patients and to and to study its possible relation to patient’s glycemic control. METHODS: An epidemiological, cross sectional, naturalistic study was carried out in Spanish Primary Care centres. Patients >18 years with diabetes mellitus type 1 (T1DM) or type 2, with insulin treatment (T2DM-i) or not (T2DM-n.i), were enrolled in the study (consecutive cases sampling). The last value of glycosylated hemoglobin (HbA1c) of each patient, reported in 2006, defined the glycemic control as satisfactory (HbA1c ≤ 7%) or unsatisfactory (HbA1c>7%). Treatment compliance was obtained from Morisky Green questionnaire (MG) and illness knowledge from 6 questions about diabetes. RESULTS: A total of 679 patients were enrolled in the study: 52.4% female; age 65.2 (13.7); BMI 28.81 (4.66); type of diabetes: 11.5% T1DM, 26.2% T2DM-i and 62.3% T2DM-n.i; mean time from diagnoses 11.9 (9.25) years. 53% of patients achieved satisfactory control (T1DM: 29.5%, T2DM-i: 31.5% and T2DM-n.i: 63.8%; p<0.001). A total of 67.4% of patients presented a good compliance according to MG scores; no significant differences were found between glycemic control groups or type of diabetic patients. 99.0% of patients knew they had diabetes or that they were taking drugs for it, 33.2% knew what HbA1c means, 22.7% knew their last HbA1c value and 21.4% knew the appropriate HbA1c value; unsatisfactory glycemic control group demonstrated better knowledge in relation to these 3 questions (p<0.05). 81.6% of diabetic patients with high blood pressure (N=516) and 79.9% of patients with high cholesterol levels (N=488) were aware of having these levels above appropriate values; no significant differences were found between glycemic control groups or type of diabetes in these two questions. CONCLUSIONS: This study shows that percentage of patients with a good compliance and knowledge about their illness should increase. More efforts should be done in these areas to improve treatment effectiveness.