Preclinical models and methods aimed at detecting and quantitating minimal residual disease (MRD) after autologous bone marrow transplantation (BMT) for acute myeloid leukemia (AML) could facilitate assessment of innovative therapeutic strategies for their antileukemic potential. Among the various techniques exploited to identify MRD, polymerase chain reaction (PCR) proved to be a valuable tool in instances in which clonogeneic markers are involved during the evolution of disease. In human AML, however, detection of MRD by PCR is limited to a minority of subgroups, as clonospecific markers are absent or presently unknown. Although gene labeling has proved to be efficient in detecting marker-devoid leukemia cells in preclinical models, detection and quantitation by PCR have not yet been considered. We therefore developed an experimental model in which detection and quantitation of genetically marked murine AML cells are based on a highly sensitive two-step nested PCR and competitive PCR protocol, respectively. We further demonstrated its applicability to a murine syngeneic BMT model that was designed to monitor minimal numbers of gene-tagged AML cells at various time intervals after transplantation. Our results showed that detection and quantitation could reproducibly be achieved at levels as low as one in 10(6) and 10(5) cells, respectively.