Production and purification of hydrophobin HFBII has recently been the subject of intensive research, but the yield of production needs to be further improved for a generic use of this molecule at industrial scale. In a first step, the influence of different carbon sources on the growth of Trichoderma reesei and the production of HFBII was investigated. The optimum productivity was obtained by using 40. g/L lactose. Carbon starvation and excretion of extracellular enzyme were determined as two main conditions for the production of HFBII. In the second phase, and according to the physiological mechanisms observed during the screening phase, a bioreactor set up has been designed and two modes of cultures have been investigated, i.e. the classical submerged fermentation and a fungal biofilm reactor. In this last set-up, the broth is continuously recirculated on a metal packing exhibiting a high specific surface. In this case, the fungal biomass was mainly attached to the metal packing, leading to a simplification of downstream processing scheme. More importantly, the HFBII concentration increased up to 48.6. ±. 6.2. mg/L which was 1.8 times higher in this reactor configuration and faster than the submerged culture. X-ray tomography analysis shows that the biofilm overgrowth occurs when successive cultures are performed on the same packing. However, this phenomenon has no significant influence on the yield of HFBII, suggesting that this process could be operated in continuous mode. Protein hydrolysis during stationary phase was observed by MALDI-TOF analysis according to the removal of the last amino acid from the structure of HFBII after 48. h from the beginning of fermentation in biofilm reactor. Hopefully this modification does not lead to alternation of the main physicochemical properties of HFBII. \textcopyright 2014 Elsevier B.V.