In a context of the potentially epidemic nature of both diabetes mellitus and depression, and the negative effects reported in cases of comorbidity, this review suggests that the association of the two conditions is multifaceted. Increased risks of prevalent depression and incident depression among diabetic patients have been reported in community studies. Even more consistent is the finding supporting psychosomatic hypotheses regarding the increased risk of diabetes among depressed patients. A recent relevant finding is the increased risk of diabetes reported in depression that is commonly found in the community, namely nonsevere, persistent, untreated depression. In view of the negative implications of the comorbidity of depression and diabetes, the suggestion that all clinically relevant cases of depression found in the community should be treated seems logical. However, new studies seem mandatory to document the efficacy of treatment of depression and the safety of antidepressant use in cases of comorbidity.