BACKGROUND: Type 2 diabetes (T2DM) doubles the odds of comorbid depression. Depression is a strong predictor of developing T2DM. The aim of the study was to compare depressed patients with T2DM to non-depressed ones with respect to demographic, psycho-social, clinical, anthropometric and metabolic characteristics; to examine the relationship between glycemic control and depression severity in depressed patients; to estimate the risk factors of depression. SUBJECTS AND METHODS: A group of depressed diabetic patients comprising those with a Major depressive episode, first or repeated (ICD-10; 1992) and endocrinologist-diagnosed T2DM, duration ≥5 years on oral, insulin therapy or both (N=46) and non-depressed ones (N=44) (90 in total) of both genders (<65 years) were included in this cross-sectional study. Laboratory and non-laboratory measures were performed.. The patient Health Questionnaire (PHQ-9) and a structured interview (MINI) were used to establish diagnosis, while the Beck Depression Inventory (BDI; cut off ≥16) was used to assess the severity ofdepression. Scaling of Life Events (SLE) for self-assessment of life events and Problem in Areas in Diabetes (PAID) for self-assessment of diabetes distress were also performed. RESULTS: Statistically significant higher rates of psychiatric heredity, neuropathy, higher level of diabetes related distress and a greater number of life events in depressed patients compared to non-depressed ones were found. There was a statistically significant positive correlation between BDI somatic subscore and the HbA1c level (r=0.343; p=0.020). The level of diabetes related distress (OR=1.084; p=0.000), total number of life events (OR=4.528; p=0.001) and neuropathy (OR=8.699; p=0.039) were statistically significant predictors of depression using logistic regression. CONCLUSIONS: The results obtained showed that depression in diabetic patients was predicted by both psychological (diabetes related distress, life events) and disease-specific variables (neuropathy). The severity of self-reported somatic depressive symptoms significantly correlated with the HbA1c level in depressed diabetic patients.