Boom sprayer is widely used in large farm crops because of its high working efficiency and favorable spraying effect. But there are still some problems in cotton defoliant spraying in Xinjiang, China. Cotton is planted in a high density in Xinjiang, the row space is (10+66) cm, leaves in two adjacent rows are seriously overlapped, the lower leavers are poorly sprayed, so the defoliation effect is poor, and the cotton quality is degraded. To solve this problem and improve the defoliant droplets coverage on the cotton canopy, the original boom spraying was modified, and the spraying pardameters was optimized by the central combination test and design concept of Box-Behnken based on a single-factor test. A quadratic polynomial model of droplets coverage was created by using working parameters including horizontal spraying boom height, hang boom height and nozzles angle as the influential factors and the mean droplets coverage on cotton canopy as the target function, and the effectiveness of mode and interaction of factors were analyzed. The model was optimized and analyzed using the regression analysis method and response surface analysis method of software Design-Expert 7.0.0, and the optimal combination of spraying parameters was obtained. The results showed that the droplets coverage on cotton canopy were influenced by boom height, sprayer height and angled nozzles sequentially from large to small, and the optimal combination of spraying parameters was under horizontal spraying boom height of 134 cm, hang spraying boom height of 27.5 cm and nozzles angle of 21°. The mean droplets coverage of experimental value and predicted value on cotton canopy were 19.6% and 20.43% respectively in such conditions, and the relative error to the estimated value on the model was -4.25%. The research result can provide a reference for further optimizing the spraying parameters of cotton defoliant sprayer. [ABSTRACT FROM AUTHOR]