The development of acetyl-cholinesterase inhibitors, and the prospect of future therapies to prevent, or modify, the course of Alzheimer's disease necessitates greater accuracy in diagnosis of this heterogeneous disease. Current diagnosis is based on clinical criteria and neuropathology. This is not always sufficient, and the development of sensitive and specific biomarkers would enable earlier and more accurate diagnosis. Genetic markers, such as Apolipoprotein E4, and cerebrospinal fluid markers such as beta-amyloid and tau, support a diagnosis of Alzheimer's disease. The latter can also predict conversion from mild cognitive impairment to dementia. Imaging markers improve diagnostic accuracy by reflecting brain function or aspects of in vivo pathological changes. In order for such biomarkers to become clinically useful, however, effective treatments need to become available, and long-term follow-up studies are necessary to evaluate the relevance of cross-sectional biomarker changes for the longitudinal natural history of the disease.