Introduction: Gene therapy offers long term solutions for chronic diseases, whereby the transgene is continuously expressed upon single vector administration. However in some cases it would be desirable to tightly regulate or switch off transgene expression. Methods: We are investigating regulated gene expression based on the mifepristone (MFP)-inducible GeneSwitch system. The GeneSwitch protein comprises yeast Gal4 DNA-binding domain, a human p65 activation domain and a MFP controlled domain derived from the human progesterone receptor. The classical GeneSwitch system consists of two expression cassettes on two separate vectors; one containing the GeneSwitch sequence and one containing the transgene. We compared this two-vector system to a single-vector system, where the two cassettes were put into one vector for efficacy in vitro and in vivo. Results: We show inducible expression of EPO, IGF and GNDF obtained in vitro upon addition of MFP to cells transfected with plasmids containing GeneSwitch and the gene expression cassette. The kinetics of EPO mRNA and protein expression followed a dose dependent fashion in the range of 0.1 to 10 nM MFP and reached a plateau at higher MFP concentrations. Surprisingly, the GeneSwitch protein expression decreased 48h after MFP induction. The indicibility of the single versus the two-vector system of GeneSwitch-EPO was compared. Both systems were equally inducible based on total amount of EPO produced in the presence of MFP and related to background expression in the absence of MFP. In vivo proof of concept was obtained for EPO in the liver. EPO is characterized by clear expression kinetics in plasma and raises blood hematocrit, hence provides a reliable in-life read-out for gene inducibility. Mice were injected with different doses of AAV5-AAT-GeneSwitch-EPO and gene expression was induced in two separate rounds at 4 and 8 weeks p.i. EPO plasma levels increased approximately 2-logs in the single or two-vector system-injected mice, compared to un-induced groups. Moreover in the absence of MFP background expression of EPO was lower in the single-vector system and hematocrit levels were unaffected. Measurements of MFP in tissue matrices and in plasma by mass spectrometry show the presence of MFP in plasma and liver, validating applicability of the GeneSwitch system in the liver. Conclusion: Overall, our data indicate that transgene expression can be repeatedly regulated in the liver using the GeneSwitch system and provides us with a novel AAV5 vector for further development.