The sensitivity of assays designed to monitor minimal residual disease (MRD) by RT-PCR in leukemia depend on quality and quantity of RNA derived from peripheral blood (PB) and bone marrow (BM) leukocytes. Shipment of material may lead to RNA degradation resulting in a loss of sensitivity and, potentially, false negative results. Furthermore, degradation may lead to inaccurate estimates of MRD in positive specimens. We sought to determine feasibility and efficacy of a novel blood collection and processing system which is based on integrated RNA stabilization at the time of phlebotomy (PAXgene Blood RNA Kit) by comparison with standard methods of RNA extraction (cesium chloride gradient ultracentrifugation and RNeasy Mini Kit) using unstabilized EDTA anticoagulated PB. In 26 patients with chronic myelogenous leukemia (CML) on therapy, PB was processed after a storage time at room temperature of 2 and 72 h according to these protocols. BCR-ABL, total ABL and glucose-6-phosphate dehydrogenase (G6PD) mRNA transcripts of PB samples were quantified as a measure for response to therapy and RNA integrity. RNA yield expressed as the ratio of ABL transcripts after a storage time of 72 h/ABL transcripts after a storage time of 2 h at room temperature was significantly higher with the stabilizing method (median 0.40) compared to the RNeasy method using unstabilized PB (median 0.13, P = 0.01). Furthermore, ratios BCR-ABL/ABL after 72 vs 2 h still correlated well using the PAXgene method (r = 0.99, P < 0.0001) in contrast to the standard method which did not (r = 0.65, P = 0.03). Even investigation of complete cytogenetic responders with very low tumor burden showed a good correlation of ratios BCR-ABL/ABL compared to the reference method. Comparable results were achieved using G6PD transcripts as standard. We conclude that the new PAXgene stabilization method could improve RNA quality and the comparability of molecular monitoring within and between multicenter trials.