BACKGROUND: The health burden associated with comorbid depression and diabetes in older community residents in middle income countries is unclear. METHODS: Data came from a statewide representative sample (N = 6963, age ≥ 60) in Brazil. Controlled polytomous logistic regression was used to determine whether four mutually exclusive groups (all possible combinations of the presence or absence of depression and diabetes) differed in sociodemographic characteristics, social resources, health behaviors, and selected health conditions. RESULTS: While 2.37% were expected to have comorbid depression/diabetes given sample base rates (depression: 20.92% [1457/6963]; diabetes: 11.35% [790/6959]), comorbidity was present in 3.62% (52.5% beyond expectation; P<0.0001; OR = 1.58, 95% Confidence Interval 1.29-1.95). Depression without diabetes was reported by 17.3%, and diabetes without depression by 7.7%. In controlled analyses, the depression group had poorer socioeconomic status and health behaviors, and a greater likelihood of vascular, respiratory, and musculoskeletal problems than the diabetes group. Vascular, respiratory, and urinary problems were exacerbated in comorbid depression/diabetes; the comorbid group was also more likely to be female and younger. LIMITATIONS: Cross-sectional design. CONCLUSIONS: To our knowledge, this is the first study that explicitly reports on all four possible depression/diabetes combinations in an older representative community-resident sample, using controlled analyses to identify unique associations with sociodemographic characteristics and other health conditions. The burden of comorbid depression/diabetes in Brazil, a middle income country, appears to be comparable to that found in higher income countries. So, similarly, depression without diabetes had a greater odds of adverse sociodemographic and health conditions than diabetes without depression; comorbid depression/diabetes was more likely in women and young elderly, and the odds of vascular, respiratory, and urinary conditions was increased significantly. Attention to comorbid depression/diabetes as a unique entity is needed.