Alzheimer's disease (AD) has been linked to a state of cerebral and systemic inflammation. The objective of the present study was to determine whether singular markers or a set of inflammatory biomarkers in peripheral blood allow discrimination between AD patients and healthy controls at the individual level. Methods: Using bead based multiplexed sandwich immunoassays, 25 inflammatory biomarkers were measured in 164 serum samples from individuals with early AD and age-matched cognitively healthy elderly controls. The data set was randomly split into a training set for feature selection and classification training and a test set for class prediction of blinded samples (1 : 1 ratio) to evaluate the chosen predictors and parameters. Multivariate data analysis was performed with use of a support vector machine (SVM). Results: After selection of sTNF-R1 as most discriminative parameter in the training set, the application of SVM to the independent test dataset resulted in a 90.0% correct classification for individual AD and control subjects. Conclusions: We identified sTNF-R1 from a marker set consisting of 25 inflammatory biomarkers, which allowed SVM-based discrimination of AD patients from healthy controls on a single-subject classification level with a clinically relevant accuracy and validity. Although larger sample populations will be needed to confirm this diagnostic accuracy, our study suggests that sTNF-R1 in serum-either as singular marker or incorporated into a biomarker panel-could be a powerful new biomarker for detection of AD. In addition, selective inhibition of TNF-R1 function may represent a new therapeutic approach in AD.