Statistics based experimental designs were used to optimize the medium for antifungal active substances production from a newly isolated Paenibacillus polymyxa Cp-S316 in shaker flask cultivation. The medium components having significant effect on the production were first identified using a fractional factorial design. Then steepest ascent method was employed to approach the experimental design space. followed by an application of response surface methodology for further optimization. A quadratic model was found to fit the antifungal active substances production. Response surface analysis revealed that the optimum values of the tested variables for the production of active substances were 12.3 (g/l) lactose, 17.5 (g/l) peptone, 0.4 (g/l) sodium nitrate, 4.5 (g/l) magnesium sulfate and 100 (g/l) potato. A production of 4687.71 mu g/ml, which was in agreement with the prediction, was observed in verification experiment. In comparison to the production of basal medium, 3.05-fold increase had been obtained. (c) 2008 Elsevier Ltd. All rights reserved.