The optimization of nutrient levels for the production of thermostable beta-amylase by Clostridium thermosulfurogenes SV2 in solid-state fermentation (SSF) was carried out using response surface methodology (RSM) based on the central composite rotatable design (CCRD). The design contains a total of 54 experimental trials with the first 32 organized in a fractional factorial design and experimental trials from 33 to 40 and 51 to 54 involving the replications of the central points. The design was employed by selecting magnesium chloride, potato starch, ferrous sulphate, pearl millet flour and corn steep liquor as model factors. Among the five independent variables studied, except pearl millet flour, all the nutrients were found significant. 16.5% potato starch, 2.5% corn steep liquor, 0.015% ferrous sulphate and 0.1% magnesium chloride have been found optimal for the production of thermostable beta-amylase. Strain SV2 produced 36% more beta-amylase in the nutritionally optimized SSF medium. (C) 2002 Elsevier Ltd. All rights reserved.