Suitable reference genes for correct quantification of reverse transcription PCR (qRT-PCR) have to be constantly expressed in all samples under investigation. Thus, it is mandatory to determine expression stability of control genes before normalization. We aimed to establish optimum inducing concentrations for the prototypical enzyme and drug transporter inducer rifampicin in LS180 cells and concurrently assessed reference gene stability under rifampicin treatment. LS180 cells were treated with increasing concentrations of rifampicin (up to 200 μM), and expression of eight different reference genes and some target genes (CYP3A4, ABCB1, and ABCC1) was quantified using real-time qRT-PCR. To check whether the results can be generalized, HepG2 cells were also investigated. We demonstrated that higher concentrations of rifampicin (>50 μM) change the expression of reference genes and thus may complicate and adulterate normalization of qRT-PCR data. The results stress the need for proper validation of potential reference genes in respective cells, tissues, and particular experimental conditions. Programs like geNorm and NormFinder alone do not warrant an adequate choice of the most suitable reference gene. Scrutiny of the reference gene expression and plausibility of the data remain necessary and protect from erroneous quantification and misinterpretation of qRT-PCR data.