Compensation of matrix effects is the main obstacle to achieve accurate quantification in LC-MS/MS. Therefore, quantitative LC-MS/MS typically relies on addition of internal standards (ISs). Here, we present a systematic comparison of ISs from a routine LC-MS/MS method for bile acid (BA) analysis with a focus on tauro-conjugated. Both human serum based quality control (QC) and patient samples were quantified with a variety of stable isotope labeled (SIL) BAs including D5-tauro BAs. As expected, SIL-ISs that match the endogenous analytes provide the highest data quality (precision, accuracy). We could not observe systematic correlation of data quality with chemical similarity or proximity in retention time of ISs to the analyte. However, both accuracy and precision of QCs and patient's concentrations showed significant correlation. This provides evidence that calculation of matrix-based QCs with various ISs could be applied for the selection of ISs whenever matching SILs are not available. Moreover, data calculated without ISs exhibited a poor data quality for both QCs and patients' concentrations. Considering these results, data generated without ISs should be interpreted with great care and may not be suitable for proposal of biomarkers unless solid quantitative data could confirm initial findings.