Factorial design and response surface analyses were used to optimize the production of inulinase (2,1-beta-D-fructan fructanohydrolase, EC 3.2.1.7) by Kluyveromyces marxianus ATCC 16045, using sucrose as carbon source. Effects of aeration, agitation and type of impeller (disk turbine, marine, pitched blade) were studied in a batch stirred reactor. Two factorial designs 2(2) were carried out. Agitation speed varied from 50 to 550 rpm (revolution per minute), aeration rate from 0.5 to 2.0 vvm (air volume/broth volume-minute). It has been shown that the enzyme production was strongly influenced by mixing conditions, while aeration rate was shown to be less significant. Additionally, the increase in the agitation speed is limited by the death rate, which increases drastically at high speeds, lowering the enzyme production. Also, the impeller type has significant influence in the production, the disk impeller at 450 rpm and aeration at 1.0 vvm led to an activity of 121 UI/mL, while the pitched blade was shown to be the best impeller for this process, leading to the best production, 176 UI/mL, at 450 rpm and 1.0 vvm. The maximum shear stress for inulinase production was about 0.22 Pa, since higher values cause higher cell death rates, affecting the enzyme production. The same results were confirmed with another microorganism, which was also sensible to shear stress. Therefore, it has been concluded that in some cases, mainly when the microorganism is sensible to shear stress, the interaction between mass transfer and mechanical stress should be considered in scale up processes. (c) 2004 Elsevier Inc. All rights reserved.