Today, the most reliable diagnosis for Alzheimer's disease (AD) is the post mortem identification of amyloid plaques, consisting of the Amyloid-beta (Abeta) peptide, (and neurofibrillary tangles) in the brain of the patient. Great efforts are being made to identify reliable biomarkers for AD that are suitable for minimal invasive early diagnosis and prognosis of AD. During the past years, body fluids of AD patients were assayed for their content of total or soluble Abeta(1-40) or Abeta(1-42) concentrations using classical (ELISA) or non-classical (with additional signal amplification) read-out. Cerebrospinal fluid (CSF) concentrations of soluble Abeta(1-42) are reduced by 40 to 50 % in AD patients compared to age-matched healthy controls as confirmed in more than 30 studies, with both sensitivity and specificity exceeding 80 % in most of the studies. Thus, it was suggested that low levels of CSF Abeta(1-42) might be useful for preclinical diagnosis. Because the current average sensitivity of AD biomarker detection in the CSF is approximately 85 %, these assays do not offer a considerable increase in predictive value over existing algorithms based on neuropsychological and imaging modalities. Regarding the amyloid cascade hypothesis, Abeta oligomers and aggregates are directly involved in the pathogenic process. Therefore, presence of Abeta aggregates seem to be the most direct disease biomarker for AD and increasing effort is being made into the development of methods suitable for the detection of different Abeta aggregates in body fluids like CSF and plasma. We therefore give an overview of the current state of Abeta aggregate specific detection.