The optimization of nutrient levels for the production of alpha-amylase by Bacillus amyloliquefaciens was carried out using response surface methodology (RSM) based on the 2(3) factorial central composite design (CCD). This procedure limited the number of actual experiments performed while allowing for possible interactions between three components. RSM was adopted to derive a statistical model for the effect of starch, peptone and yeast extract (YE) on alpha-amylase production. The P-value of the coefficient for linear effects of starch and YE concentration was < 0.0001, suggesting that this was the principal experimental variable, having the greatest effect on the production of alpha-amylase. The optimal combinations of media constituents for maximum alpha-amylase production were determined as 12.61 g L-1 starch, 2.83 g L-1 peptone and 1.25 g L-1 YE. The optimization of the medium resulted not only in a 34% higher enzyme activity than unoptimized medium but also in a reduced amount of the required medium constituents. (c) 2006 Society of Chemical Industry.