Quantitative reverse transcription PCR (RT-PCR) has become an important tool for studying functional gene expression. However the most often used cycle threshold (CT)-based method, primarily related to the required amplification efficiency determination via serial dilution, can call into question the level of quantitative reliability and accuracy that can be achieved, in addition to the impracticalities inherent to CT-based methodologies. In this study, an alternative method, named the sigmoidal curve-fitting (SCF) method, was compared with the classic CT method for two target genes (XRCC4 and HIF-1alpha) and a reference gene (HPRT). The PCR conditions were optimized for each gene on a LightCycler apparatus. Fluorescence data were fitted to a four-parametric sigmoidal function, and the initial messenger RNA (mRNA) copy number was determined by a theoretical fluorescence (F0) value calculated from each fitting curve. The relative expression of the target gene versus that of the reference gene was calculated using an equation based upon these F0 values. The results show that the F0 value had a good linearity with the initial number of target genes between 10(7) and 10(1) copies. The reproducibility tests showed that the variations of initial target quantity were well reflected by F0 values. Relative expression of target gene calculated by the SCF method and by the CT method showed similar results. In our hands, the SCF method gave reliable results and a more precise error description of quantitative RT-PCR.