This chapter asks how molecular models “re-present” and “stand in” as proxies for protein molecules. It examines the use of mimetic, three-dimensional models in the history of the life sciences, comparing and contrasting these to models of protein structures. This chapter seeks to trouble standard accounts of realism and representation, proposing the concept of rendering as a performative theory that can grapple with the ways that molecular models not only re-present the molecular world, but also materialize some aspects of living matter, if not others. It documents the anxieties expressed by practitioners when their models of otherwise imperceptible and indeterminate phenomena are taken as realistic representations of concrete objects. This chapter argues that in spite of the representational failings of their models, practitioners find ways to keep open the space between the model and its referent, and that this practice allows them to situate the limits of their knowledge.