Abstract: Seabuckthorn is one of the most important ecological cash crops in modern agriculture. The existing seabuckthorn forests in China are accounted for more than 90% of the world's total crop area. However, the current manual harvesting cannot fully meet the large-scale production in recent years, because the fruits can be clustered in the thorny branches. It is also a high demand for seabuckthorn harvesting machinery rather than manual ones. In this study, the movable double blades cutting device was developed to investigate the effect of the cutting parameters on the harvesting performance of seabuckthorn branches. The optimal combination of cutting parameters was also achieved after the experiment and simulation. The sample branches were four years old, 300 mm in length, with a moisture content of 43.13% to 53.83% (average 47%). The test instruments included the self-developed double blades cutting test device, a video recorder, and a digital camera. The cutting test device consisted of three units (the cutting, branch feeding, and control unit). The single factor and orthogonal tests were also carried out to verify the simulation. The experimental factors in the single factor tests were the average cutting speed (0.4-0.8 m/s), sliding cutting angle of the blade (7.5°, 15°, and 22.5°), blade height (35, and 70 mm), and the number of the movable blade (single, double), where the average cutting speed was set at 5 levels. A regression model was established for the orthogonal tests. The Box-Behnken method was designed with a three-factor, three-level test scheme. The average cutting speed (0.4-0.6 m/s), branch feeding speed (0.5-0.7 m/s), and sliding cutting angle of the blade (7.5°, 15°, and 22.5°) were selected as the experimental factors, whereas, the peak cutting force per unit diameter and cutting power consumption per unit area were set as the evaluation indexes. Response surface analysis was used to optimize the regression model. The experimental results show that the evaluation indexes decreased with the increase in average cutting speed, sliding cutting angle of blade, and blade height. The values of these indexes were smaller for the double movable blades, compared with the single ones. The average cutting speed and sliding cutting angle of blade presented the highly significant effects on the evaluation indexes ( P<0.01), while the branch feeding speed shared a non-significant effect on the evaluation indexes ( P>0.05). The key factor for the quality of the cutting section was the average cutting speed when the branch feeding speed was determined. The optimal combination of cutting parameters for the double movable blades was: the average cutting speed of 0.45 m/s, the branch feeding speed of 0.64 m/s, and the sliding cutting angle of 9.4°. Specifically, the peak force per unit diameter was 53.33 N/mm, and the cutting power consumption per unit area was 69.87 kJ/m² under the combination. Three groups of verification tests were carried out for the experimental optimization. The relative error of the regression model was less than 5%, indicating the reliable optimization of cutting parameters. Therefore, the optimized conditions were reasonable in this case. The optimized parameters (such as the average cutting speed, branch feeding speed, and sliding cutting angle of the blade) can be expected to effectively reduce the peak force per unit diameter for the cutting power consumption saving per unit area. The finding can provide the data support for the subsequent development of seabuckthorn harvesting equipment.