Abstract: Abstract: In order to find out the safe buffer areas of pesticide drift during the aerial spraying by N-3 unmanned aerial vehicle (N-3 UAV), researches on pesticide drift were done, by simulations and experiments as flight speed was 3 m/s, crosswind velocities were 1m/s, 2m/s and 3m/s, flight heights were 5 m, 6 m and 7 m. According to the parameters of the N-3 UAV and the aerial spraying operations, the whole equipment, atomizing surface and the entire computational domain were grid-processed first. A turbulence model by means of approximate solutions of N-S equations with appropriate boundary condition was developed and the two-phase flow of rotor wind field and pesticide-spraying was simulated. In the simulations, the equations were discretized by second-order upwind format based on the finite volume methods, the fluxes were calculated by the ROE scheme. Iterative calculations were processed using Gauss - Seidel method to analyze the droplets flow rate changes under the wind generated by the UAV's rotor until to the flow stability when the calculation converges residuals dropped more than three orders of magnitude, then liquid concentration and density of each grid were achieved, resulting droplet drift and deposition. And the model was verified by experiments. A solution of tracer (Rhodamine-B) mixed with water in a certain concentration was selected to replace the pesticide for aerial spraying, to monitor the wind velocity as well as the temperature and humidity. The N-3 UAV flight heights were changed in the experiments. The target area was a rectangle by 50 m×20 m, and the fly route was the centerline. Mylar cards with a diameter of 90 mm every 2 m downwind, 5 m along the flight direction out of the target area were sampled to collect the droplets drifted, forming a 50m×10m drift sample area. The drift amount at each sampling position was measured by means of a fluorescence spectrophotometer to obtain the regularities of droplets drifted by the wind. The simulations showed that the drift distance downwind (along positive Z-axis) and the maximum amount of drift position rear the UAV (along negative X-axis) were not significant with 5 m, 6 m, 7 m flight height under the same crosswind velocity, while which were relatively significant as crosswind velocity changed from 1 m/s to 3 m/s with the same flight height. The comparison results of amounts and trends of droplet drift (when flight height was 6 m, crosswind velocity was 1 m/s, 2 m/s, 3 m/s) showed that the simulated and measured curves were coincident, and the correlation determination (R2) were 0.7068, 0.8451 and 0.6859 respectively, which showed that the research had a significance for determining the buffers before aerial spraying to insure the safety. Also, the following conclusions can be drawn by the simulation and experiment results: the crosswind velocity is greater than the aerial spraying height as the droplet drift affecting factors; the droplet drift only occurs downwind of the spraying field, and as the crosswind velocity is 1-3 m, 8-10 m buffer zones should be considered downwind the spraying field for safe aerial spraying.