The downwash airflow of a multi-rotor unmanned aerial vehicle (UAV) in hover can improve the penetration rate of droplets in the canopy of fruit trees, but the uncertain distribution of the downwash airflow can also lead to the generation of uneven distribution of droplets. With a co-axial eight-rotor plant protection UAV as the research object, the computational fluid dynamics (CFD) method is adopted in this study to establish the coupled flow field model of the downwash airflow and the two-phase flow model of airflow-droplets in hover. The spatial and temporal distribution law of the downwash airflow is analyzed, and the impact of the motion law of the droplets in the spray field is investigated. The results show that the downwash airflow shifts from an unsteady state to a steady state over time. Various wind vortices such as wake vortex, ground diffusion airflow, and vortex ring are generated during the evolution process, with a “negative velocity channel” of downwash airflow velocity identified in the central region under the rotors. The farther the airflow is from the rotor, the larger the area field of wake vortex airflow. Consequently, there is a gradual coupling of the left and right adjacent wake vortices from mutual independence, with the “spiral effect” of the downwash wind field offset by the mutual coupling of the upper and lower rotors of the UAV. When the installation position of the nozzle falls within the below-rotor region, the distribution of droplet deposition becomes more penetrative than the central axis installation position of the nozzle. It indicates that the below-rotor installation position of the nozzle can prevent droplets from entering into the coupling region of the trailing vortex, which reduces droplet drift. According to the field test, the measurement and simulation results of the downwash air velocity are consistent, with the minimum and maximum relative error reaching 4% and 29%.