Abstract: Mixed flow pumps have been widely used in agricultural irrigation and drainage, industrial water circulation, and ship propulsion systems, due to their excellent overall performance. It is a high demand to optimize mixed-flow pumps for better energy conversion efficiency in recent years. This study aims to explore the influence of geometric and loading parameters on the energy characteristics, and then to further improve the optimization upper limit of the guide vane mixed flow pump. A parametric optimization was also carried out on the impeller of a guide vane mixed-flow pump with a specific speed of 511. Among them, the geometric and the loading parameters were taken as the design parameters, whereas, the pump section efficiencies at 0.85Qdes and 1.15Qdes were taken as the optimization objectives, and the pump section head under the design condition was as the constraint condition. Taguchi design and numerical simulation were also combined in this case. The inverse design method was adopted to verify the accuracy of numerical simulation. The results show that only 27 schemes needed to be constructed in the optimal design of 13 design parameters with 2 optimization objectives and 1 constraint using orthogonal design. Therefore, Taguchi design performed better in the multi-factor coupling optimization, which effectively reduced the amount of calculation. The extreme difference analysis showed that the effect of each parameter on the efficiency at 0.85Qdes was ranked in the descending order of Kh, Ls, NDs, LEs, NCs, ST, NCh, Lh, Ks, Ts, NDh, LEh, and Th (LE is the leading edge loading, NC is the horizontal coordinates of first loading point, ND is the horizontal coordinates of second loading point, K is the slope of middle straight, Lh and LS represent the X axis horizontal coordinates of the intersection point of the blade leading edge with hub and shroud, Th and TS represent the X axis horizontal coordinates of the intersection point of the blade trailing edge with hub and shroud, the subscripts h and s represent hubs and shroud, respectively.), in order to maximize the efficiency at 0.85Qdes in the levels of 1, 3, 2, 1, 3, 2, 3, 2, and 1, respectively. Similarly, the effect of each parameter on the efficiency at 1.15Qdes was ranked in the descending order of the LEs, Kh, LEh, Lh, Ls, Ks, NDh, Ts, NCs, NCh, NDs, ST, and Th, in order to maximize the efficiency at 1.15Qdes with the levels of 3, 1, 3, 3, 3, 1, 1, 3, 1, 3, 2, 2, and 2, respectively. The effect of each parameter on the head at 1.0Qdes was ranked in the order of NDh, Kh, NCs, ST, Ks, NDs, LEs, Ts, Ls, Lh, NCh, LEh, and Th. According to the influence of each parameter on the optimization objectives and constraint, the geometric parameters (Lh and Ls) and the load parameters (Kh, LEs, NCs and Ks) posed a significant impact on the performance of the mixed-flow pump, which should be considered in the optimization design. Compared with the original, the pump section head of the optimal model was basically unchanged at 1.0Qdes, which fully met the constraint requirements. Meanwhile, the pump section efficiencies at 0.85Qdes and 1.15Qdes increased by 0.90 and 2.25 percentage points, respectively, which fully meeting the optimization requirements. The internal flow analysis showed that the pressure and velocity distribution near the blade leading edge of the optimized model was significantly improved, compared with the original. In addition, the hydraulic losses of downstream components were also significantly reduced in the optimized model, which was mainly due to the improvement of the uniformity of flow field distribution at the outlet of the impeller. In conclusion, this finding can provide an important reference for the parameterized optimization of turbomachinery, in order to save computational resources and maximize the optimization effect.