Resin-derived contaminants added to samples during column chemistry are shown to cause matrix effects that lead to inaccuracy in multi-collector inductively coupled plasma mass spectrometry measurement of small natural variations in Cd and Zn isotopic compositions. These matrix effects were evaluated by comparing pure Cd and Zn standards and standards doped with bulk column blank from the anion exchange chromatography procedure. Doped standards exhibit signal enhancements (Cd, Ag, Zn and Cu), instrumental mass bias changes and inaccurate isotopic compositions relative to undoped standards, all of which are attributed to the combined presence of resin-derived organics and inorganics. The matrix effect associated with the inorganic component of the column blanks was evaluated separately by doping standards with metals at the trace levels detected in the column blanks. Mass bias effects introduced by the inorganic column blank matrix are smaller than for the bulk column blank matrix but can still lead to significant changes in ion signal intensity, instrumental mass bias and isotopic ratios. Chemical treatment with refluxed HNO(3) or HClO(4)/HNO(3) removes resin-derived organic components resulting in matrix effects similar in magnitude to those associated with the inorganic component of the column blank. Mass bias correction using combined external normalization-SSB does not correct for these matrix effects because the instrumental mass biases experienced by Cd and Zn are decoupled from those of Ag and Cu, respectively. Our results demonstrate that ion exchange chromatography and associated resin-derived contaminants can be a source of error in MC-ICP-MS measurement of heavy stable element isotopic compositions.