This review addresses recent developments in amyloid beta (Abeta), total tau (t-tau), and phosporylated tau (p-tau) protein analysis, in cerebrospinal fluid (CSF) and plasma as biomarkers for dementia. Recent research focused on the protection of patients with mild cognitive impairment (MCI) into dementia and the differential diagnosis of Alzheimer's Disease (AD). A combination of Abeta42 and t-tau in CSF can discriminate between patients with stable MCI and patients with progressive MCI into AD or other types of dementia with a sufficient sensitivity and specificity. Regression analyses demonstrated that pathological CSF (with decreased Abeta42 and and increased tau levels) is a very strong predictor for the progression of MCI into AD. Furthermore, CSF measurements of p-tau and Abeta42 can assist in diagnosing vascular dementia or frontotemporal dementia in the differential diagnosis of AD indicated by a reasonable sensitivity and specificity. Whether tau in combination with Abeta42 or in combination with the Abeta37/Abeta42 or Abeta38/Abeta42 ratio aids in the discrimination between AD and Lewy Body dementia remains to be elucidated. Cross-sectional research could not demonstrate significant differences for Abeta40 and Abeta42 in plasma between AD and controls. However, a recently published longitudinal study showed high baseline Abeta40 levels, especially when combined with low baseline Abeta 42 levels, are a strong risk factor for the development of dementia. This emphasizes the importance of performing longitudinal studies in addition to cross-sectional ones. The origin of plasma Abeta and its transport between CSF and plasma, however, needs further clarification. In conclusion, progress has been made regarding Abeta and tau as biomarkers for dementia, both for differentiation between stable MCI and progressive MCI patients and for the differential diagnosis of AD. Future research should aim to validate these recently published results, preferably in pathologically confirmed AD patients. In addition, it is important to standardise research in terms of study design (longitudinal, minimal follow-up period of 5 years), type of researched parameters ( total or p-tau, type of Abeta peptides), type of matrix (CSF and plasma) and data analysis (establishment of predefined cut-off values, type of ratio, type of marker combination).