DNA topoisomerase I (Topo I) is involved in DNA replication, transcription, recombination and repair. Clinical interest has focused on Topo I as it is the molecular target of camptothecin (CPT), used in first and second lines of treatment for different cancer types. Furthermore, it is well demonstrated that the patients who best responded to CPT-based chemotherapy were generally those with the greatest tumoral Topo I expression and/or activity. We developed a sensitive, simple and reproducible method to measure Topo I mRNA expression in human cancer samples. Experiments were performed in two steps. First, we checked the accuracy of the reverse transcription-polymerase chain reaction (RT-PCR) method by testing intra- and interassay reproducibility of Topo I and G6PDH gene amplification in different cell types. We observed that crossing-points (Cps) were different, depending on the cell type, dilution or cDNA concentration, but that the intra- and interassay Cp standard deviation (SD) never exceeded 0.77% and 1.39% for Topo I amplification, or 1.63% and 2.9% for G6PDH amplification, respectively. Secondly, we used our method to measure Topo I mRNA levels in primary tumor samples obtained from 27 patients with advanced colorectal cancer and 10 patients with pharyngeal/laryngeal cancer. The accuracy of G6PDH as a housekeeping gene was tested by analyzing its correlation with the mRNA level of a second housekeeping gene, porphobilinogen deaminase ( PBG-D ) in the tumoral samples. We found that the normalized Topo I/G6PDH mRNA ratios were significantly correlated with that of Topo I/PBGD in colorectal tumors (r 2 =0.47, p=0.02) but not in pharyngeal/laryngeal tumors (r 2 =0.35, p=0.3). Neither ratio showed any significant association with clinicopathological parameters, such as gender, age, tumor size, or grade and lymph node status. We believe that RT-PCR is a reliable and highly reproducible technique. However, the choice of the reference gene is an important point and must be defined based on the samples studied.