This paper describes a global optimization of a recirculation flow type casing treatment in centrifugal compressors of turbochargers. The global optimization for the recirculating flow type casing treatment has been performed based on the existing casing treatment. The optimization approach to the recirculation flow type casing treatment for the centrifugal compressor, incorporating meta-model assisted evolutionary algorithm, computational fluid dynamics analysis technique, artificial neural network, and genetic algorithms has been presented. For the baseline design of the casing treatment, numerical approach is validated with an experimental result from a test rig. The technical issue of the casing treatment is found to be the dropping of the adiabatic efficiency at smaller flow rate condition. In this study, the objective of the optimization is to improve adiabatic efficiency under multipoint mass flow rate conditions. The shape of the casing treatment has been parameterized by six parameters. The numerical optimization result gives the optimized recirculation casing shape, which has a possibility to improve the efficiency not only at design flow rate but also at smaller flow rate. The improvement in adiabatic efficiency at off-design point is discussed by means of the improved flow incidence at the inlet of the rotating impeller. The influence on adiabatic efficiencies at both design and off-design conditions is discussed by the sensitivity of the recirculation flow rate. It is found that the optimized design of the casing treatment provides optimized recirculation flow rate.