Background: Diabetes is a risk factor for dementia,but the issue whether this concerns only vascular dementia or also Alzheimer’s disease is debated. We compared the clinical diagnoses and abnormalities on brain MRI in patients with or without diabetes who received standardised, detailed diagnostic studies at a memory clinic, in order to establish whether one specific type of dementia or specific MRI abnormalities were more common in diabetes. Patients and methods: Patients who visited our memory clinic between January 2002 and June 2004 were divided into a group with (n = 42) or without diabetes (n = 389). The diagnoses were recorded, and MRI scans were rated for (sub)cortical atrophy, medial temporal lobe atrophy, infarctions, and white matter changes. Results: The proportion of Alzheimer’s disease (36%versus 28%; OR 1.1 (95%CI 0.5-2.2), adjusted for age and sex), vascular dementia (5%versus 2%; OR 2.4 (0.5-12.1)), and so called “cognitive impairment no dementia”(24%versus 17%; 1.3 (0.6-2.9)) was similar in patients with or without diabetes. On MRI lacunar and cortical infarctions were more common and cortical atrophy more pronounced among diabetic patients. By contrast, the severity of white matter changes was similar in the two groups. Conclusion: The relative frequency of different diagnoses among diabetic and non-diabetic patients attending a memory clinic was similar, indicating that diabetes does not predispose to one particular subtype of dementia. The imaging findings support the notion that the increased risk of cognitive decline and dementia in elderly subjects with diabetes is due to dual pathology, involving both cerebrovascular disease and cortical atrophy. Background: Diabetes is a risk factor for dementia, but the issue whether this concerns only vascular dementia or also Alzheimer’s disease is debated. We compared the clinical diagnosis and abnormalities on brain MRI in patients with or without diabetes who received standardized, detailed diagnostic studies at a memory clinic, in order to establish whether one specific type of dementia or specific MRI abnormalities were more common in diabetes. The diagnoses were recorded, and MRI scans were rated for (sub) cortical atrophy, medial temporal lobe atrophy, infarctions, and white matter changes. Results: The proportion of Alzheimer’s disease (36% versus 28%; OR 1.1 (95% CI 0.5-2.2), adjusted for age and sex, vascular dementia (5% versus 2%; OR 2.4 (0.5-12.1)), and so called “cognitive impairment no dementia” % versus 17 On MRI lacunar and cortical infarctions were more common and cortical atrophy more pronounced among diabetic patients. By contrast, the severity of white matter changes was similar in the two groups. Conclusion: The relative frequency of different diagnoses among diabetic and non-diabetic patients attending a memory clinic was similar, indicating that diabetes does not predispose to one particular subtype of dementia. The imaging findings support the notion that the increased risk of cognitive decline and dementia in elderly subjects with diabetes is due to dual pathology, involving both cerebrovascular disease and cortical atrophy.