Background. Due to certain limitations, the bioprocess development for protease production needs more convenient and realistic statistical approach instead of conventional optimization technique. For an economic bioprocess with enhanced protease yield, Response Surface Methodology (RSM) based on Central Composite Design (CCD) was employed and evaluated in this study. Materials and methods. The fermentation was performed with a mutant strain, Bacillus licheniformis MZK05M9 (BlM9) using molasses, urea and CaCl 2 .2H 2 O as carbon, nitrogen and trace element sources respectively in shake flask. The conditions for fermentation were maintained with temperature, pH and agitation at 37°C, 7.5 and 150 rpm respectively. The required number of trials were determined by investigating each variable (Molasses, Urea and CaCl 2 ) at five levels: -α, -1, 0, +1 and +α through CCD with protease yield as the response function and the interaction effects as well as optimal parameters were obtained by using Minitab software. The significance of the independent variables and their interactions were tested by means of analysis of variance (ANOVA) with a 95% confidence level and 3-D surface plots were developed through RSM. Results. Upon 20 trials, the optimum values of the tested variables for maximum alkaline protease production as predicted through CCD and RSM were as 0.63%, 0.16%, and 0.11% (w/v) for Molasses, Urea and CaCl 2 .2H 2 O, respectively. The protease activity in Conventionally Optimized (CO) medium was 410 U/ ml and it was predicted as 463.1 U/ ml for statistically optimized medium. Upon experiments with the optimized medium, the protease activity was estimated as 560 U/ ml which was 36.6% (i.e. 1.36 fold) higher than that of CO medium. Conclusion. The efficiency of the enzyme in solubilizing the whole feathers was also assessed which indicated that the enzyme produced with cheap substrates could be utilized as a cost effective and eco-friendly agent in poultry feed formulation, leather processing etc.