The sequence of a protein normally determines which amino acid residues will form a helices, and which one b sheets, to an extent that allows secondary structure prediction to be made with a reasonable reliability. Nevertheless, non-native helical structures are observed during in vitro folding of several model proteins and may even occur during protein biosynthesis within the ribosomal exit tunnel. Moreover, non-native b sheet structures are common in amyloid fibrils formed by a variety of pathogenic and even non-pathogenic proteins and peptides. In all of these cases, the formation of a helix precedes the appearance of b sheet, which suggests that conversion from the simpler, more local helix structure to the often more convoluted sheet architecture during folding and pathogenic misfolding processes could be a unifying principle of general importance. A better understanding of this switching process, and the ability to design molecular systems which can be induced to switch between these conformations will have a significant impact on fields ranging from fundamental biochemistry through to applied technology and medicine.