Antigen-specific immune responses in multiple sclerosis have been studied for decades, but the target antigens of the putatively autoaggressive B and T cells still remain elusive. Here, we summarize recent strategies which are based on the direct analysis of biopsy or autopsy specimens from patients. Since this material is extremely scarce, the experimental methods need to be exceptionally sensitive. We describe technologies to distinguish (auto) aggressive T cells from irrelevant bystander lymphocytes by analyzing clonal expansions in relation to the morphological location of the cells in the tissue lesions. We then discuss approaches to clone matching alpha- and beta-chains of the antigen-specific T cell receptor (TCR) molecules from single T cells. This is necessary because usually, several clones are expanded and are diluted by many irrelevant cells. The matching TCR chains from individual T cells can be resurrected in hybridoma cells which may then be used for antigen searches. We discuss strategies to identify antigens of gammadelta- and alphabeta-TCR molecules, such as biochemical methods, candidate antigens, human leukocyte antigen requirements, synthetic peptide, and cDNA libraries. These strategies are tailored to characterize the antigens of the membrane-anchored, low-affinity TCR molecules. The strategies to identify (auto) reactive B cells or immunoglobulin (Ig) molecules are fundamentally different, because Ig molecules are water-soluble and have high affinities. We further discuss proteome-based approaches, techniques that analyze Ig-chains from single B cells, and a repertoire-based method that compares Ig-proteomes and Ig-transcriptomes. The first method detects Ig antigens directly, whereas the latter two methods allow reconstruction of Ig molecules, which can be used for antigen searches.