The objective of this study was to identify through quantum mechanical quantitative structure activity relationships (Q-QSARs) chemical structures in dental monomers that influence their mutagenicity. AMPAC, a semiempirical computer program that provides quantum mechanical information for chemical structures, was applied to three series of reference chemicals: a set of methacrylates, a set of aromatic and a set of aliphatic epoxy compounds. QSAR models were developed using this chemical information together with mutagenicity data (Salmonella TA 100, Ames Test). CODESSA, a QSAR program that calculates quantum chemical descriptors from information generated by AMPAC and statistically matches these descriptors with observed biological properties was used. QSARs were developed which had r2 values exceeding 0.90 for each study series. These QSARs were used to accurately predict the mutagenicity of BISGMA. a monomer commonly used in dentistry, and two epoxy monomers with developing use in dentistry, GY-281 and UVR-6105. The Q-QSAR quantum mechanical descriptors correctly predicted the level of mutagenicity for all three compounds. The descriptors in the correlation equation pointed to components of structure that may contribute to mutagenesis. The QSARs also provided 'dose windows' for testing mutagenicity, circumventing the need for extensive dose exploration in the laboratory. The Q-QSAR method promises an approach for biomaterials scientists to predict and avoid mutagenicity from the chemicals used in new biomaterial designs.