[Aim] Influences of different survey methods on community structure and biodiversity in transgenic cotton fields were determined to define the most effective survey method for arthropod community research.[Method] In 2016 and 2017, we investigated the arthropod communities in two transgenic cotton (GGK2, N15-5) fields and their parental non-transgenic (K312, J14) equivalents using three survey methods:(1) investigate every 7 days during the whole growth period (early May-mid to end of September); (2) investigate in the critical growth periods only (mid May, mid June, mid July, mid August, mid September); (3) investigate three times during the peak of cotton bollworm Helicoverpa armigera occurrence from mid June to late August. We calculated species richness, Shannon-Wiener diversity index, evenness index and Simpson dominance index.[Result] We found significant differences in the total number of species and species richness surveyed by three methods with the maximum value by the first and the minimum value by the third method. In 2016, the total number of species and of arthropod community and the pest-sub community in herbicide-resistant cotton (GGK2) fields decreased significantly compared with its parental non-transgenic (K312) field surveyed by all three methods, but there was no significant difference in other indicators. In addition, there was no obvious difference on the Shannon-Wiener diversity index, evenness index, and Simpson index in the same cotton field between 2016 and 2017.[Conclusion] The survey method had influence on the total number and the abundance of species, and had no significant effect on the Shannon-Wiener diversity index, evenness index and Simpson dominance index. Therefore, we should choose the appropriate survey method according to different research purposes.