In this paper, we conducted a statistical optimization of the medium components for the production of rifamycin SV by Amycolatopsis mediterranei MM2. In order to maximize the yield of rifamycin SV, a Plakett-Burman experimental design (PBD) was initially utilized in the screening of the medium components among 11 nutrients. Glycerol and yeast extract were determined to influence significantly the yield of rifamycin SV. Then, a central-composite experimental design (CCD) was utilized in order to optimize the concentrations of the screened components accqired using the PBD and to predict the mutual interactions occurring between the screened components. The predicted optimal glycerol and yeast extract concentrations were determined to be 43.8 and 9.5g/L, respectively. At this optimum point, the predicted rifamycin SV yield was 490.5 mg/L, whereas the corresponding experimental yield was $480.3{pm}43.8 mg/L$.