The aim of this study was to investigate the relationship between total homocysteine levels in people with Type 2 diabetes and cognitive status. Fifty patients from our diabetes unit (30 females/20 males) with diabetes were enrolled. All patients had fasting blood samples taken for measurement of cardiovascular risk factors; total cholesterol and triglyceride concentrations and other lipid fractions (lipoprotein (a), low density lipoprotein (LDL-cholesterol), high density lipoprotein (HDL-cholesterol)), glucose, HbA(1c) and homocysteine. 24-h urine collection was used to measure creatinine clearance and microalbuminuria. Vitamin B-12 and folate were measured to assess vitamin status. All diabetic patients were assessed for late complications and a Mini-Mental State Examination (MMSE) was performed. The patients were 64.6 (49-78) years old with body mass index (BMI) of 29.6 +/- 6.3 kg/m(2), and duration of diabetes of 8.9 +/- 6.7 years. A univariant correlation analysis was performed among cardiovascular risk factors and vitamins with total MMSE score. Total homocysteine was inverse by correlated with MMSE score (r=-0.38; P<0.05) of the other measures of cardiovascular risk, microalbuminuria showed an inverse correlation with MMSE score (r=-0.51:P<0.01). Lipoproteins, glucose control and vitamin status were not correlated MMSE score. In the multiple regression model only microalbuminuria remained in the model, showing a decrease of one point in the MMSE result with each milligram of microalbuminuria, adjusted for confounding factors. Cognitive status in type 2 diabetic was correlated with homocysteine levels and microalbuminuria, this last endothelial damage marker remaining as an independent risk factor of cognitive deterioration.