Abstract: In order to obtain the optimal combination of bacterial strains for the fermentation production of ethanol from Jerusalem artichoke (Helianthus tuberosus), the recombinant bacterial strain R32 was selected at first to optimize its inulinase producing conditions. The highest inulinase activity of the R32 was 298.8 U·mL^-1, and the best culture medium formulae were 1% yeast powder (w/v), 2% peptone (w/v), and 0.5% glycerol (v/v) for YPG culture medium, and 1% yeast powder (w/v), 2% peptone (w/v), and 1% methanol (v/v) for YPM culture medium. The medium pH was the natural initial pH. Then, the Saccharomyces cerevisiae (S.c) and Kluyveromyces (Klu) were selected to compare their ethanol producing capability when fermented separately or in admixture under the conditions of adding R32 crude inulinase broth or not. The results showed that the ethanol production reached the maximum of 11.37% after S.c and Klu being onestep fermented in admixture for 84 h without adding R32 crude inulinase broth, and reached the maximum of 11.43% after S.c and Klu being twostep fermented in admixture for 72 h with the addition of R32 crude inulinase broth. The maximum ethanol yields under the twostep fermentation of S.c and Klu were almost the same. Because of the lower cost and more simplicity of operation, onestep fermentation was more appropriate for the industrial production of ethanol, though it took a longer time. After optimization, the production of ethanol from Jerusalem artichoke by S.c and Klu mixed in onestep fermentation was up to 11.82% under the conditions of 225 g·L^-1 of inulin, 40 g·L^-1 of urea, 15% of inoculation rate, and pH 5.