Lignocellulosic biomass, including wood, can be transformed into bioethanol using steam explosion as pretreatment to improve saccharification and fermentation steps. Pretreatment is however the most expensive part of the process in terms of CAPEX and OPEX and requires to be optimized. In order to evaluate the link between pretreatment efficiency and cost, three contrasted wood species (oak, poplar and spruce) were pretreated with continuous steam explosion at pilot-scale following full factorial designs. Response surfaces obtained were combined with an economic assessment to determine trade-offs aiming at maximizing both fermentable sugars released during the enzymatic hydrolysis step and bioethanol yield during the fermentation step as well as minimizing costs of pretreatment in an industrial context. Results showed that bioethanol yields were highly dependent on wood species and that high severities of pretreatment were not the most relevant to apply. Optimal conditions of pretreatment corresponding to 70 % and 48 % of bioethanol producible from oak and poplar, respectively, were defined. The desirability function that has been modelled thus helps designing adapted pretreatment conditions regarding bioethanol production and process cost.