The use of human cytochrome P450 (CYP) enzymes is increasing for the production of drug metabolites used for drug safety testing and doping analysis. Major challenges are high-priced cofactors, poor stability, and comparatively low activities. We have shown previously that production of specific metabolites in milligrams to gram scale is feasible using human CYPs recombinantly expressed in fission yeast. In this study, we sought to improve the activities of human CYP3A enzymes by genetic engineering. Two side chains (Pro293 and Arg409) of known activating human CYP3A polymorphic variants were-separately or together-introduced into the wild-type forms of each of the three enzymes CYP3A4, CYP3A5, and CYP3A7, respectively. Different effects of the two mutations and their combination on enzyme activity were monitored using both polar and nonpolar substrates. Interestingly, the CYP3A7 double mutant displayed a strong increase in activity with respect to testosterone 6 beta-hydroxylation (300 % of wild-type activity) and luciferin-6'-pentafluoro-benzyl ether turnover (400 % compared to wild type), while the single mutant CYP3A5(Pro293) showed 370 and 400 % of wild-type activity towards 6 beta-hydroxylation of testosterone and 16 alpha-hydroxylation of dehydroepiandrosterone, respectively. Overall, six out of seven newly created mutants displayed increased activity with at least one of the tested substrates. These results support the notion that pharmacogenetic knowledge can directly contribute to the improvement of biotechnological processes.