Amyloid fibres are stable, persistent and highly ordered aggregates of mis-folded protein that accumulate in tissues and are a prominent feature of the pathology of a wide range of human diseases. The presumed role of amyloid as a causative factor of tissue damage is based largely on 'guilt by association'. However, growing understanding of the nature of amyloid, its formation by a nucleated growth mechanism from destabilised and partially unfolded precursors and its persistence at sites of deposition has provided the foundation for the development of approaches to inhibit amyloid formation and enable its clearance. In spite of intensive study, our understanding of the detailed structure of amyloid itself remains incomplete although 'crossed-beta' structure is clearly a common constituent. On the other hand detailed structural understanding of transthyretin, beta-secretase and serum amyloid P component is contributing to the design of small molecule compounds to target amyloid. Thyroxin mimetics stabilise the native tetrameric protein structure, beta-secretase inhibitors will limit the production of the amyloidogenic Abeta1-42 polypeptide. Compounds that crosslink serum amyloid P component rapidly deplete the plasma and amyloid-bound pool of this protein. The efficacy of these compounds as drugs to prevent formation or enable removal of amyloid will provide a stringent test of the 'amyloid hypothesis' of disease.