The diagnosis of Alzheimer's disease (AD) is mainly performed by excluding other disorders with similar clinical features. In addition, an analysis of symptoms and signs, blood studies and brain imaging are major ingredients of the clinical diagnostic work-up. However, the diagnosis based on these instruments is unsatisfactory, indicating the need of a highly sensitive and reliable approaches, selective for AD and based on biological markers. Ideally, such markers should reflect the pathophysiological mechanisms of AD, which according to the current hypotheses, derive from the actions of two major protein aggregates, the extracellular beta-amyloid (Abeta) plaques and the neurofibrillary tangles. Since AD is a multifactorial disease, other factors that cause neuronal insult and that contribute to neuronal degeneration in AD include free radical and oxidative stress promoting molecules, proinflammatory cytokines and neurotoxic agents. In this context, the search for anomalous levels or changes in the molecular patterns of Abeta(1-42) or Abeta(1-40), hyperphosphorylated tau isoforms, oxidation products in the cell or cytokines such as interleukin-1 or 6 facilitates the selection of biomarkers in AD. There is clear evidence that the cerebrospinal fluid (CSF) levels of beta(1-42) are significantly reduced in AD patients as compared with senile controls, while increased levels of tau have been revealed. The CSF levels of these proteins reflect their metabolism in the central nervous system. Approaches using ELISA and immunochemical methods for the quantification of these markers in CSF have been preferentially used. Diagnosis criteria and number of patients exhibits variations in the different reports, while clinico-pathological studies are scarce. An increasing number of studies suggest that supplementary use of these CSF markers preferably in combination, adds to the accuracy of an AD diagnosis.