Pseudomonas elongata, a kappa-carrageenase producing marine bacterium, was isolated from West coast of Gujarat, India and for the first time, its nutritional requirement for kappa-carrageenase production was defined using statistical optimization method. The important medium components identified by Plackett-Burman design were kappa-carrageenan, yeast extract, NaCl, K2HPO4 and KH2PO4. These variables were further used in central composite design where carrageenase production, pH and biomass were responses used in different models to evaluate individually fit ones. Quadratic model was found to be fit for carrageenase production and pH, however, for biomass, 2F1 model was found to be fit after performing power transformation. In multiresponse analysis, medium formulation consisting of (g l(-1)): kappa-carrageenan 0.2, yeast extract 0.4, NaCl 10.8, K2HPO4 0.3 and KH2PO4 0.1, was found to predict kappa-carrageenase production of 32.08 units mg(-1) protein with overall highest desirability of 0.679 as compared to other formulations. The selection of model was done on basis of high adjusted R-2 value and lowered p-value for each model in individual analysis of each response. Through desirability analysis, it was found that for carrageenase production by P. elongata, biomass played an important role as compared to pH. Through statistical optimization method, 32-fold increase in kappa-carrageenase production was achieved as compared to medium reported in the literature. Furthermore, through response surface experimental design, medium consisting of (g l(-1)): kappa-carrageenan 0.2, yeast extract 0.7, NaCl 22.2, K2HPO4 0.1 and KH2PO4 0.2, was found to produce an activity of 60.7 units mg(-1) protein. However, this does Dot depict the interactions among all the components as well as their combined effect on the responses. (c) 2006 Elsevier Inc. All rights reserved.