A large part of the reliability of reverse transcription quantitative polymerase chain reaction (RT-qPCR) data depends on technical variations. Such variations are mainly attributable to the reverse transcription step. Standardization is a key factor in decreasing the intersample variability. However, an ideal standardization is not always possible, and compromises must be found. Due to technical requirements, the current consensus is that a constant amount of total RNA should be used for the RT step (CA-RT). Because RNA isolation yields are variable, such a practice requires the use of variable volumes of nucleic acid extracts in RT reaction. We demonstrate that some RNA extracts contain both exogenous and endogenous inhibitors. These inhibitors induce a decrease in RT efficiency that significantly impairs the reliability of RT-qPCR data. Conversely, these inhibitors have a slight effect on the qPCR step. To overcome such drawbacks, we proposed to carry out the RT reaction with a constant volume of RNA extract by preserving a constant RNA amount through the supplementation of yeast transfer RNA (CV-RT). We show that CV-RT, compared with the usual CA-RT, allows us to decrease the RT-qPCR variability induced by intersample differences. Such a decrease is a prerequisite for the reliability of messenger RNA quantification.