The production of a highly alkaline and thermostable pectinase of Bacillus pumilus was optimized in submerged fermentation using Plackett-Burman design and response surface methodology. Three fermentation variables (C:N ratio, K2HPO4, and pH), which were identified to significantly affect pectinase production by Plackett-Burman design were further optimized using response surface methodology of central composite design (CCD). An over all 34- and 41-fold increase in enzyme production was achieved in shake flasks and lab fermenter by the optimization of variables using statistical approaches, respectively. The enzyme was optimally active at pH 10.5 and 50 degrees C, and selectively degraded only the noncellulosic gummy material of ramie (Boehmeria nivea) fibres causing 10.96% fibre weight loss, and therefore, the enzyme could find application in fibre processing industry. The use of the enzyme in fibre processing reduces the use of alkali, and the associated alkalinization of water bodies. (c) 2005 Elsevier Ltd. All rights reserved.