Quality of a performance has relationship with operational process of a product or whenever the company really implements or conducts a service and measurement towards the degree of level of satisfaction for the consumer. In this research the writer tries to model the optimization of the performance of general purpose manganese battery using polynomial regression meta-model approach with the surface responses of the methodology from some factors influencing the quality of manganese battery. There are some basic considerations that underlie the research; one of them is the company has not known exactly the most optimum performance condition from the general purpose battery type towards the influence of storage time and temperature. By implementing the polynomial regression meta-model with Response Surface Methodology (RSM), we can model a optimization solution to the combination of input variable of temperature and storage time at certain observation area by estimating the optimum output value (response value) so that we can obtain the most optimum battery performance in order to meet the consumers demand. There are matters to be considered in implementing RSM: level of confidence (a), meta-model fitting area, step measurement on the steepest ascent and central composite design. The result of the research on general purpose manganese battery shows the mathematical model of the optimization of performance of general purpose manganese battery using appropriate polynomial regression meta-model of Y (T, S) = 62.385 + 1.282 T + 0.00029 TS - 0.201 T2 - 0.0052 S2 using variable combination to the influence of temperature 32.347°C, where the storage time of 63.306 days obtain the optimum battery performance of 103.663 minutes, using temperature performance index (PI) of 32°C (rounded) and the storage time of 90 days obtain the performance index-1 (PI-1) of 127.53% and PI-2 of 112.82%. Where the initial condition of temperature Performance Index of 20°C with the storage time of 90 days obtain PI-1 of 124.34% and PI-2of 109.81%. This shows that there is improvement and increase of battery performance of 3% for PI-1 and PI-2.