[Objectives]  To determine the effects of different sampling methods on the structural characteristics of arthropod communities in soybean fields. [Methods]  We investigated and compared the pros and cons of the net capture (WB), vacuum-suction (XC), fixed-point plant (DD), yellow trapping (HP), and trap-cup trapping (XJ) methods. [Results]  The WB method collected the most taxa, but there was no significant difference in number of taxa collected between this and the XC method. The DD method collected the least taxa. The WB and XC methods collected the most families, species and individuals. The DD method collected the most Thysanoptera, the HP method collected the most Hymenoptera, Diptera, Coleoptera, Acarina and Collembola, and the XJ method collected the most Orthoptera, Hymenoptera, Coleoptera and Collembola. In terms of community structure, the results obtained using the WB and XC methods were 0.75 similar; similarities among other methods were all less than 0.42. Combinations of two methods such as, WB+DD, WB+HP, and WB+XJ, obtained community characteristic indices > 0.75 similar to those obtained by the combination WB+XC+DD+HP+XJ. Each of three combinations containing the WB method; WB+DD+HP, WB+DD+XJ, and WB+HP+XJ, obtained a community characteristic index > 0.85 similar to that obtained by WB+XC+DD+HP+XJ. Specific combinations of methods can be selected according to the taxa of interest. [Conclusion]  Correct choice of combinations of different sampling methods can improve the quality of data obtained while reducing the workload. The survey sampling method can be selected to suit the objectives of the study.