The time-optimal control of MMA-MA copolymerization in a CSTR is treated for a grade-change operation. The control objective is to bring system specifications, such as polymer production rate and polymer composition, to desired values in minimum time. The initiator concentration is chosen as a manipulated variable. Two cases are considered : One is to get the time optimal control for both a desired polymer production rate and a desired weight composition of MA in dead copolymer; the other is only for a desired polymer production rate. In numerical calculation, the final t_f⁽ⁱ⁾ is chosen at each iteration as the time at which the system reached steady-state. This allows us to use a more easily manipulatable form of the performance index and at the same time not directly use the free final time boundary condition.