The impacts of two treatments on the natural enemy community diversity was evaluated through field test. The results showed that the number of family and population of natural enemies in the plots using buprofezin and chlorantraniliprole (treatment Ⅰ) was larger than that using imidacloprid and chlorpyrifos (treatment Ⅱ). The Shannon-Wiener diversity index and evenness index of the plots with treatment Ⅰ were significant higher than those with treatment Ⅱ. The relative dominance of Lycosidae and Tetragnathidae in the plots with treatment Ⅰ was significantly higher that with treatment Ⅱ and that in the control plots. The relative dominance of Braconidae and Linyphiidae was significantly lower with treatment Ⅰ than that with treatment Ⅱ. The results suggested that using buprofezin and chlorantraniliprole to control rice planthopper and Cnaphalocrocis medinalis will be safer than using imidacloprid and chlorpyrifos.