Response surface methodology (RSM) based on central composite design was successfully applied to redesign MRS media for maximizing both biomass and bacteriocin production from Lactobacillus plantarum NH40. First, glucose and yeast extract were chosen as the best carbon and nitrogen sources based on classical optimization results of one factor at time which also revealed the possibility of eliminating peptone and meat extract from the original composition of medium without affecting the growth and bacteriocin production. Statistical experimental design based on a regression model generated using the Design expert 7 software showed that the optimum concentrations of glucose, yeast extract, tween80, NH₄Cr, CH₃COONa and K₂PO₄ were 40, 19.9, 1, 3.06, 7, 1.25 g/L respectively for maximum production of biomass (15.87 mg/mL) and bacteriocin (634.74 U/mL). In addition, from the analysis of variance, yeast extract with F-value 77.2 and glucose with 185.4 were the most effective factors on biomass and bacteriocin production. Formulation of empirical model explained that the interaction among factors showed that the determination coefficient R² of biomass and bacteriocin production were 0.8777 and 0.8539 respectively. Furthermore, the accuracy of model of the optimized MRS medium suggested by design expert 7 for both biomass and bacteriocin was verified and results showed that concentrations of biomass and bacteriocin were 15 mg/mL and 640AU/mL respectively, which were approximately closed to predicted values.