Abstract: A tea vibrating-shifting machine is one of the most important equipment in the tea refining process. The low screening performance of traditional vibrating-shifting machines cannot fully meet the large-scale production in recent years, due to the low sieve net rate and high error screen rate. In this study, the virtual prototype was constructed to combine with the screening principle of the vibrating-shifting machine by Solidworks. A discrete element simulation model of tea particles was established by EDEM software. A systematic analysis was made to clarify the motion state of tea particles on the sifting surface. The dynamic simulation was carried out on the vibrating-shifting process of tea particles. A single-factor test was also conducted to obtain the average velocity and average interaction force curve of tea particles. The simulation results show that the connecting rod length, the crank radius, and the sift inclination of the surface were the main influencing factors in the screening performance of the tea vibrating-shifting machine. A three-factor three-level quadratic rotation orthogonal test was designed to perform the regression analysis on the test data using Design-Expert software, where the connecting rod length, the crank radius, and the sift inclination of the surface were taken as the influencing factors, while the sieve net rate and the error screen rate were as the evaluation indexes. The regression analysis showed that the best screening performance of the tea vibrating-shifting machine was achieved, where the sieving rate of the tea vibrating-shifting machine was 94.5%, and the false sifting rate was 4.61%. An optimal combination of parameters was the connecting rod length of the tea vibrating-shifting machine was 1 977 mm, the crank radius was 25 mm, and the sift inclination of the surface was 2.8°. The verification test was carried out with the optimal structural parameters of the tea vibrating-shifting machine. The false sifting rate of the tea vibrating-shifting machine before optimization was 90.7%, the false sifting rate was 5.12%, and the production efficiency was 293 kg/h. After optimization, the false sifting rate of the tea vibrating-shifting machine was 93.8%, and thefalse sifting rate was 4.73%. The verification test and regression analysis were basically the same agreement. At the same time, the production efficiency of the tea vibrating-shifting machine was 319 kg/h. The false sifting rate increased by 3.42%, the false sifting rate decreased by 7.62%, and the production efficiency increased by 8.87%, compared with before optimization. This finding can provide a theoretical basis for the optimization of tea screening equipment.