BACKGROUND/OBJECTIVES:It is unknown if a specific fatty-acid composition influences the development of Alzheimer's disease (AD). Nutrition is a possible target for prevention of dementia and especially omega-3-based fatty acids (n-3 FAs) have previously been suggested to be beneficial for cognition. The objective was to ascertain whether serum FAs predicts the risk of incident AD and dementia in a longitudinal population-based cohort. Keywords: fatty acid; dementia; Alzheimer's disease INTRODUCTION Prevention of dementia, and particularly of Alzheimer's disease (AD), is a major challenge for researchers and clinicians. As treatment options for dementia still are very limited, knowledge of whether the risk of dementia can be reduced, for example by consumption of a healthy diet, is extremely important for decision makers and for the general public.The composition of fatty acids (FAs) in the neuronal membranes reflects the diet in animal studies.1 Observational studies have related low fish intake and low blood levels of omega-3 FAs (n-3 FAs) to increased risk of dementia and AD.2 However, results from post-mortem studies on FAs in the brain and clinical trials are less conclusive. 3,4 As dementia itself affects dietary habits, a longitudinal prospective study design is more suitable for investigating the relationships than cross-sectional epidemiological studies. Further, biomarkers such as serum esterified FAs are most probably more objective and accurate in evaluating the true dietary fat quality than for example, food frequency questionnaires.