Most studies using real-time PCR are taken semiquantitatively and assume a steady level of expression forthe so-called housekeeping genes. By absolute real-time PCR we demonstrate that the transcript amounts of two of the most popular internall controls (coding GAPDH and beta-actin) fluctuate dramatically across diverse mouse or human tissues. This raises the question about the inaccuracy of these genes a squantitative references in tissue-specific mRNA profiling. Target genes chosen for absolute real-time PCR analysis are involved in DNA repair, regulation of gene expression, and oxidative stress response. Hence, they code for 8-oxoG-DNA glycosylase/AP-lyase, major AP-endonuclease, and heme oxygenase-1. Quantitations reported: i) determine mouse-to-mouse variability in basal gene expression, ii) establish organ- and embryo-associated differences in mouse, iii) compare mouse and human tissue-specific profiles, iv) examine the time course (30-240 min) expression in liver and lung of mice treated with paraquat (superoxide generator) at 30 mg kg(-1) (one half LD50 value), and v) explore the utility of absolute real-time PCR in field studies with genetically diverse mice. We conclusively establish that real-time PCR is a highly sensitive and reproducible technique for absolute quantitation of transcript levels in vivo and propose its use to quantitate gene expression modulation under mild physiological exposures and for field epidemiological studies.