A statistical approach, response surface methodology (RSM), was used to study the production of extracellular protease from Bacillus sp., which has properties of immense industrial importance. The most influential parameters for protease production obtained through the method of testing the parameters one at a time were starch, soybean meal, CaCl2, agitation rate, and inoculum density. This method resulted in the production of 2543 U/mL of protease in 48 h from Bacillus sp. Based on these results, face-centered central composite design falling under RSM was employed to further enhance protease activity. The interactive effect of the most influential parameters resulted in a 1.50-fold increase in protease production, yielding 3746 U/mL in 48 h. Analysis of variance showed the adequacy of the model and verification experiments confirmed its validity. On subsequent scale-up in a 30-L bioreactor using conditions optimized through RSM, 3978 U/mL of protease was produced in 18 h. This clearly indicated that the model remained valid even on a large scale. RSM is a quick process for optimization of a large number of variables and provides profound insight into the interactive effect of various parameters involved in protease production.