Background: Diabetes and depression are common and rapidly increasing non-communicable diseases throughout the world. Currently, about 382 million people have diabetes worldwide, while depression affects approximately 350 million people. Some studies have found a frequent co-existence of depression, hyperglycemia, diabetes and diabetes-related complications. Moreover, comorbid depression in diabetes has been associated with poorer adherence to diabetes treatment regimens, increased risk of work loss and functional disability, increased mortality rates, higher health care costs, and decreased quality of life. Although the association between these two conditions has been found by several, the transcultural validity of these findings still needs to be demonstrated. Objectives: The main objectives of the study were to investigate the prevalence of type 2 diabetes and depression, and the association between depressive symptoms and newly diagnosed diabetes in Northeastern Brazil. In addition, we wanted to investigate the agreement between two different types of depression scales: the Montgomery-Åsberg Depression Rating Scale (MADRS) and the Hamilton Depression Rating Scale (HDRS). Methods: The prevalence of diabetes was assessed in the Northeast region of Brazil in a randomized population-based survey following the WHO criteria of 1999. Seven hundred and fourteen subjects participated in the study. Depression was assessed by MADRS and HDRS, before the diagnosis of diabetes was made known to the participants and investigators. Socio-demographic and economic information, as well as anthropometric measures were collected. Results: A high prevalence of diabetes was found (Total 16%; Male 13.2% and Female 17.4%). Following MADRS, the rate of depression was 15% (Male 7% and Female 19.1%). According to HDRS, the rate of depression was 15.5% (Male 8.3% and female 19.3%). The agreement between MADRS and HDRS was found to be excellent (Kappa of 0.913, p < 0.001). Depression was the second strongest risk indicator for the occurrence of diabetes after controlling for potential confounding fators. Conclusions: We found a high prevalence of both diabetes and depression in this population. Depression was a strong independent risk indicator for the occurrence of diabetes. An inverse significant association between diabetes and the risk for developing depressive symptoms was also observed. The results may indicate that the treatment of depression should be included both for prevention and treatment of diabetes.