A stable isotopically labeled (SIL) analogue is believed to be the most appropriate internal standard in a quantitative bioanalytical liquid chromatography/tandem mass spectrometry (LC/MS/MS) assay. It is assumed that a SIL internal standard always compensates for variability in chemical derivatization, sample extraction and LC/MS/MS analysis due to its nearly identical chemical and physical properties to the unlabeled analyte. Hence, the analyte to internal standard peak area ratio should be constant despite any variations in sample processing or analysis. However, in our laboratories, a deuterium labeled internal standard of carvedilol demonstrated an unexpected behavior-the analyte to internal standard peak area ratio changed with two specific lots of commercially supplied human plasma. Several experiments, including dilution of the extract with LC mobile phase and post-column infusion of the carvedilol solution followed by the injection of extracted blank plasma, have indicated that a high level of matrix suppression affected the ionization of the carvedilol-S enantiomer and its deuterated internal standard differently in these two lots of plasma. For the first time, it was clearly demonstrated that a slight difference in retention time between the analyte and the SIL internal standard, caused by deuterium isotope effect, has resulted in a different degree of ion suppression between these two analogues. This difference was significant enough to change the analyte to internal standard peak area ratio and affect the accuracy of the method.