AbstractMolecular Quantum Similarity Measures (MQSM), which allow quantitative comparison between molecular electronic density distributions, are investigated as a potential source of QSAR parameters. By computing the MQSM for all possible molecular pairs in a given molecular set, a Similarity Matrix is obtained, containing relevant information about the structural relationships within the set. Approximate Overlap‐like MQSM using several computational levels are employed to obtain similarity matrices for three molecular sets, taken as test cases: (1) non‐branched alkanes, (2) Baker triazines, (3) indole derivatives. The similarity matrices have been analyzed by means of PCA and PLS techniques. QSAR regression models were derived in order to fit available property or biological activity data. A reasonably good correlation between properties or activities and the similarity measures has been fo