Abstract  [Objectives]  To improve the efficiency of searchlight trapping for migratory pests. [Methods]  The number, species and community structure of insects trapped using four kinds of searchlight were compared, and the effect of fumigation on preventing the escape of trapped insects was determined. [Results]  A total of 234 650 insects were trapped using the four kinds of searchlight. The relationship between the number of insects trapped (Y) and the number of light sources (X) conformed to the linear equation Y = 13 398.91 + 16 459.49 X. The number of insects trapped using all 5 light sources was 3.44 times that of a single light source. During the four peak trapping periods, the number of insects trapped also increased with the number of light sources. 71 to 95 species of insect were trapped by the four kinds of searchlight. The Lepidoptera and Noctuidae were the dominant taxa. The community similarity of the insects trapped by each of the four searchlights was between medium dissimilarity and medium similarity. The community similarity of the insects trapped by each kind of searchlight during the four peak periods was medium dissimilarity and extreme dissimilarity. The number of trapped insects was 22.94 % higher than the control, and increased by 89.17 % during the peak period. [Conclusion]  Increasing the number of light sources can both significantly increase the number of migratory pests trapped using searchlights and affect the community structure of trapped insects. Adding fumigation to searchlight traps can effectively reduce the escape of trapped insects.