A　parameterization　method　is　developed　to　perform　the　multi-objective　optimization　design　of　centrifugal　impellers.　The　relationship　of　the　impeller　parameters　and　the　dimensionless　length　in　the　cylindrical　coordinate　system　is　established　by　the　transformation　of　impeller　profile　data.　Non-uniform　rational　B-spline　is　used　to　reconstruct　the　impeller　to　obtain　the　profile　of　the　blade　and　the　meridional　surface.　An　automatic　optimization　design　platform　for　the　centrifugal　impellers　is　constructed　by　the　multi-objective　genetic　algorithm　combined　with　parameterization　method　and　the　commercial　computational　fluid　dynamics　software　NUMECA.　The　Krain　high-speed　centrifugal　impeller　is　optimized　by　using　the　present　method　at　the　design　condition,　and　the　aerodynamic　performance　and　the　flow　characteristics　are　analyzed.　The　numerical　results　show　that　the　aerodynamic　performance　of　the　optimal　impellers　is　improved　to　some　extent.　The　flow　field　distribution　at　the　outlet　of　the　optimal　impellers　is　more　uniform.　The　total　pressure　ratio　and　the　isentropic　efficiency　increases　by　about　2.5%　and　1.0%　respectively.　These　findings　confirm　that　the　present　multi-objective　optimization　design　method　for　the　centrifugal　impellers　is　effective.