Arma chinensis (Fallou) as an important predatory natural enemy, has been widely used in the control of various pests. However, traditional release methods are hindered by high cost and poor timeliness, limiting the application of natural enemies. Banker plants feature the preventive introduction of natural enemies, helping to maintain natural enemy populations and achieving continuous control of pests. This study evaluated the suitability of five leguminous plants, peanut (Arachis hypogaea), soybean (Glycine max), pea (Pisum sativum), horse bean (Vicia faba), and rack bean (Phaseolus vulgaris), as banker plants for Arma chinensis in an indoor environment. The results showed significant differences (P < 0.05) in the preference rate and fecundity of Arma chinensis across different plants. The preference rate of Arma chinensis on peanut and soybean plants reached over 20%, which was significantly higher than other plants (P < 0.05). The highest average fecundity on soybean plants reached 15.2 eggs per female, accounting for 23.71% of the total eggs dropped on all plants, followed by peanut plants (14 eggs, 21.84%). No significant difference was found between soybean and peanut (P > 0.05). The hatching rate of soybean reached 93.78%, and there was no significant difference compared with horse beans or peanuts (P > 0.05). The reproductive coefficient of soybean was the highest comparing to other plants (P < 0.05). In the absence of prey, the Arma chinensis survived longest with the highest survival rates when fed on peanut and soybean leaves. A few individuals even developed into adults. In summary, soybeans and peanuts are preliminarily identified as suitable candidates for constructing Arma chinensis banker plant systems, and the effectiveness of constructing banker plant systems in the field needs further validation.