In this paper an optimization method of the runner blades in a bulb turbine based on CFD analysis is proposed.In the method the main scales of the turbine including guide vane,runner and draft are maintained.Only the runner blades are modified based on the present method.In the optimization method the runner blade is expressed by spline surface with a gather of coordinate points.The B-spline curve is used to keep the modified blades smooth.In order to make the blade optimization simple and ef- ficient,one of the coordinates is fixed and only the angles of the points are changed according to different modification purposes.Three main optimization principles based on flow diagnosis are presented here.These three principles are all based on the CFD analysis of the internal flow in bulb turbine.For the purpose of method verification,the optimization method is used in a model bulb turbine.A three dimensional steady turbulent computation is carried out through the whole passage including the bulb body,guide vanes,runner and draft tube of the bulb turbine under seven different work conditions.An SST k-ωturbulence model is used during the CFD analysis and the performance of the turbine can be achieved.The runner blade is optimized according to the three optimization principles based on flow diagnosis.The CFD analysis is conducted again on the optimized turbine and another modification is needed if the new turbine can’t satisfy the required performance.Comparison of the computational results between the original turbine and an optimized one indicates that the optimization method is practical and does improve the performance of the bulb turbine. In this paper an optimization method of the runner blades in a bulb turbine based on CFD analysis is proposed.In the method the main scales of the turbine including guide vane, runner and draft are maintained. Ofly the runner blades are modified based on the present method.In the optimization method the runner blade is expressed by spline surface with a gather of coordinate points. The B-spline curve is used to keep the modified blades smooth.In order to make the blade optimization simple and ef-ficient, one of the coordinates is fixed and only the angles of the points are changed according to the different modification purposes. Three main optimization principles based on flow diagnosis are presented here. These three principles are all based on the CFD analysis of the internal flow in bulb turbine. For the purpose of method verification, the optimization method is used in a model bulb turbine. A three-dimensional steady turbulent computation is carried out through the whole passage including the bulb body, guide vanes, runner and draft tube of the bulb turbine under seven different work conditions. An SST k-ωturbulence model is used during the CFD analysis and the performance of the turbine can be achieved. runner blade is optimized according to the three optimization principles based on flow diagnosis. CFD analysis is conducting again on the optimized turbine and another modification is needed if the new turbine can not satisfy the required performance. Comparison of the computational results between the original turbine and an optimized one indicates that the optimization method is practical and does improve the performance of the bulb turbine.