Knowledge of metabolism, enzymes so far involved, and potential enzyme-inhibiting or enzyme-inducing properties of new compounds is a key issue in drug development. Primary cultured hepatocytes, cytochrome P450 (CYP)-engineered cells and hepatoma cell lines are currently being used for this purpose, but only primary cultures can produce a metabolic profile of a drug similar to that found in vivo and can respond to inducers. Because of their limited accessibility, alternatives to replace human hepatocytes are currently being explored, including the immortalisation of hepatocytes by using different strategies (i.e. SV40 T-large antigen, conditionally immortalised hepatocytes, transfection with c-myc, cH-ras, N-ras oncogenes, transgenic animals over-expressing growth factors or oncogenes and cre-lox recombination/excision). However, none of the resulting cells has the desirable phenotypic characteristics to replace primary cultures in drug metabolisms studies. We investigated why these differentiated human hepatomas do not express CYP genes and found that the levels of certain key transcription factors clearly differ from those found in hepatocytes. It was then conceivable that re-expression of one (or more) of these transcription factors could lead to an efficient transcription of CYP genes. The feasibility of this hypothesis was demonstrated by genetic engineering of Hep G2 cells with liver-enriched transcription factors followed by the analysis of the expression of the most relevant human CYPs.