In vivo protein aggregation is strongly linked to the pathogenesis of several incurable celland neurodegenerative diseases including Alzheimer’s disease (AD) and type II diabetes (T2D). In vitro protein aggregation leads to a loss of function and complicates therapeutic application of a number of bioactive proteins or polypeptides such as insulin. Designing potent inhibitors of aggregation of such proteins is however a difficult task as these compounds should not affect protein function. Here we present a novel chemical strategy to generate inhibitors of protein aggregation that do not affect protein function by mimicking natively occuring cross-amyloid peptide interactions. We exemplify the validity of this approach by demonstrating that IAPP-GI, a designed soluble and non-amyloidogenic mimic of islet amyloid polypeptide (IAPP), which is a native insulin interaction partner and a key amyloidogenic polypeptide of T2D, is a nanomolar affinity inhibitor of non-native aggregation of insulin without affecting its biological function. As IAPP-GI also blocks with nanomolar activity cytotoxic aggregation of both IAPP and the key amyloidogenic peptide of AD Aβ, it could become a unique lead compound for the development of novel therapeutic strategies targeting both AD and T2D. In addition, our cross-amyloid-inhibitor-design approach may be applicable to other aggregation-prone polypeptides or proteins.