A new method for quantification of spectral interferences in ICP-MS, based on empirical modelling and experimental design, is presented. The relationship between interfering effect and matrix composition is deduced by performing a multiple linear regression analysis and taking into account linear and quadratic effects of matrix elements and all the potential interactions between them. In order to separate spectroscopic from non-spectroscopic interferences, analyte isotope ratio measurements were performed, except for monoisotopic elements for which the signal ratio (ratio between signal measured with and without matrix) was considered and internal standardization applied. Spectral interferences caused by sodium, calcium, chlorine and sulfur species, at the m/z of interest for the determination of several isotopes ( ⁵¹ V, ⁵² Cr, ⁵³ Cr, ⁵⁸ Ni, ⁶⁰ Ni, ⁶² Ni, ⁵⁹ Co, ⁶⁴ Zn, ⁶⁶ Zn, ⁶⁸ Zn, ⁶³ Cu, ⁶⁵ Cu, ⁷⁵ As, ¹¹¹ Cd, ¹¹⁴ Cd, ²⁰⁷ Pb and ²⁰⁸ Pb) were modelled and are discussed.