Abstract: The on-line mixing is a process with higher productivity and safer operating conditions, and can reduce waste production. In order to optimize the spraying performance of orchard air-assisted sprayer, a real-time mixing system for pesticide precision control using peristaltic pump and static mixer was developed. The components of the mixer prototype were pesticide tank, peristaltic pump, pesticide flowmeter, electromagnetic valve, water tank, filter, pressure regulator and static mixer. An electro-hydraulic proportional valve was used to continuously vary the sprayed flow rate of sprayer prototype according to a control signal depending on the canopy volume. The static mixer was used to homogenize the fluid of pesticide and water by redistributing them in the radial and tangential directions. Three insert-type configurations of static mixers named SK, SX and SD type static mixers were analyzed. Their difference was the design of stationary inserts for redistributing the fluid in the directions transverse to the main flow. With the SIMPLEC (SIMPLE-Consistent) algorithm and k?ε turbulence model, the internal flow fields in the mixing process of 3 mixers were characterized and quantified using the Fluent software for the two-phase flow mixture. The distributions of volume fraction were analyzed in detail from the cross-sections perpendicular to the main flow direction. Simulations were also conducted while varying the number of static mixing elements. Comparing the cloud map of the secondary phase volume fraction of the 3 mixers, it was found that the SX static mixer had several mixing advantages compared to other mixers, and the optimum number of elements was 5. The reason for the high productivity of SX static mixer was the multilayer design splitting the fluid in multiple layers. Taking into account the optical visibility of on-line mixing process, the glycerol solutions with carmine stain to follow the fluid motion were used as the pesticide fluids in the experiments and the SX static mixer was made of identical elements inserted in a transparent pipe. It was observed that the mixing test results of SX static mixer were in good agreement with the simulation results. It was clear that the computational fluid dynamics (CFD) analysis used in this paper could properly reveal the complicated flow characteristics in the static mixer. To assess the performance of pesticide flow control, the step rising and step falling of 84 mL/min flow rate signal sent to the peristaltic pump and the responses of the pesticide flowmeter were recorded. The time between the control signal step and the attainment of the stabilized value was approximately 0.3 s for the rising case. Similarly, the response of flowmeter required approximately 0.28 s to reach the stabilized value for the falling case. For smaller steps, the response time would be shorter. As an indication of mixing stability and mixing homogeneity, the droplets were collected and measured using the collectors, which were allowed to dry completely between successive spraying tests. The results of mixing stability test showed that the maximum error between the actual and referential concentration was less than 4.33%. The results of mixing homogeneity test showed that the coefficient of variation was no larger than 4.98%. This work can provide a reference for the design of real-time mixing mechanism and performance optimization of orchard air-assisted sprayer.