Although orphan drug applications required by the EMEA must include assessments of similarity to pre-existing products, these can be difficult to quantify. Here we illustrate a paradigm in comparing nilotinib to the prototype kinase inhibitor imatinib, and equate the degree of structural similarity to differences in properties. Nilotinib was discovered following re-engineering of imatinib, employing structural biology and medicinal chemistry strategies to optimise cellular potency and selectivity towards BCR-ABL1. Through evolving only to conserve these properties, this resulted in significant structural differences between nilotinib and imatinib, quantified by a Daylight-fingerprint-Tanimoto similarity coefficient of 0.6, with the meaning of this absolute measure being supported by an analysis of similarity distributions of similar drug-like molecules. This dissimilarity is reflected in the drugs having substantially different preclinical pharmacology and a lack of cross-intolerance in CML patients, which translates into nilotinib being an efficacious treatment for CML, with a favourable side-effect profile.