In February 1993, the group of Klaus Mosbach published their milestone study in Nature where, for the first time, non-covalent molecular imprints were employed in a competitive binding assay. In this seminal piece of work, and also for the first time, they refer to molecularly imprinted polymers as being 'antibody mimics' and hypothesised that these synthetic materials could one day provide 'a useful, general alternative to antibodies'. This perspective article examines how far we have come in the 20 years since this publication in terms of realising this hypothesis and poses the question of whether we actually need molecularly imprinted polymers to be a general alternative to antibodies.