The concept of molecular mimicry provides and elegant framework as to how cross-reactivity between antigens from a foreign agent with self proteins may trigger autoimmune diseases. While it was previously thought that sequence and structural homology between foreign and self proteins or the sharing of T cell receptor (TCR) and MHC-binding motifs are required for molecular mimicry to occur, we have shown that even completely unrelated peptide sequences may lead to cross-recognition by T cells. The use of synthetic combinatorial peptide libraries in the positional scanning format (PS-SCL) together with novel biometric prediction approaches has allowed us to describe the recognition profiles of individual autoreactive T cell clones (TCC) with unprecedented accuracy. Through studies of myelin-specific TCC as well as clones from the nervous system of patients suffering from chronic central nervous (CNS) Lyme disease it has become clear that at least some T cells are more degenerate than previously anticipated. These data will not only help us to redefine what constitutes specific T cell recognition, but also allow us to study in more detail the biological role of molecular mimicry. A recent clinical trial with an altered peptide ligand (APL) of one of the candidate myelin basic protein (MBP) epitopes in MS (amino acids 83-99) has shown that such a modified MBP peptide may not only have therapeutic efficacy, but also bears the potential to exacerbate disease. Thus, we provide firm evidence that the basic principles of cross-recognition and their pathogenetic significance are relevant in MS.