A limited number of gene expression studies have investigated the quantitative relationships between the amount of transcript, level of protein or activity/function, with disparate conclusions regarding these relationships. Collectively these studies indicate that the relevance of quantitative transcript analysis as a predictor of phenotype has to be evaluated on a gene-by-gene or even a case-by-case basis. The purpose of this study was to define a suitable marker for MDR1-dependent drug efflux, and to quantitatively investigate the relationships between the amount of transcript, protein and drug efflux in the frequently used Caco-2 cell model. The substrate specificity of digoxin, a commonly used marker for MDR1, was investigated using transgenic MDCK II or LLC-PK1 cell lines expressing the efflux proteins MDR1, BCRP and MRP2, since these proteins are localised to the apical part of the enterocyte plasma membrane and exhibit comparatively high transcript levels in the human small intestine. Relationships between levels of transcript, protein and function were investigated quantitatively using real-time RT-PCR, ECL western blot analysis and basolateral-to-apical and apical-to-basolateral efflux ratios. Our results indicate that digoxin is a specific marker for MDR1-dependent drug efflux in the Caco-2 cell drug absorption model and that MDR1 transcript abundance is at least as valid as MDR1 protein abundance as a predictor of MDR1 efflux activity.