A high intake of saturated fat and cholesterol and a low intake of polyunsaturated fatty acids have been related to an increased risk of cardiovascular disease. Cardiovascular disease has been associated with dementia. We investigated the association between fat intake and incident dementia among participants, age 55 years or older, from the population-based prospective Rotterdam Study. Food intake of 5,386 nondemented participants was assessed at baseline with a semiquantitative food-frequency questionnaire. At baseline and after an average of 2.1 years of follow-up, we screened for dementia with a three-step protocol that included a clinical examination. The risk of dementia at follow-up (RR [95% CI]) was assessed with logistic regression. After adjustment for age, sex, education, and energy intake, high intakes of the following nutrients were associated with an increased risk of dementia: total fat (RR = 2.4 [1.1-5.2]), saturated fat (RR = 1.9 [0.9-4.0]), and cholesterol (RR = 1.7 [0.9-3.2]). Dementia with a vascular component was most strongly related to total fat and saturated fat. Fish consumption, an important source of n-3 polyunsaturated fatty acids, was inversely related to incident dementia (RR = 0.4 [0.2-0.91), and in particular to Alzheimer's disease (RR = 0.3 [0.1-0.9]). This study suggests that a high saturated fat and cholesterol intake increases the risk of dementia, whereas fish consumption may decrease this risk.