<p>Multi-pond systems are a unique agricultural landscape feature in China’s rainy regions. The spatial optimization layout of multi-pond systems has the potential to process nonpoint source pollutants in agricultural watershed, yet research in this area is limited. Based on this, the methods for spatial optimizing multi-pond systems are developed to address nonpoint source pollution in agricultural watershed, and the lower reaches of the Zhegao River watershed in China are taken as study case. The research analyzes the spatial distribution characteristics of the existed ponds and the distribution of total phosphorus (TP) emission in the study area, and then, a linear optimization model is developed for multi-pond system spatial optimization configuration. The study shows that the ponds distribution is influenced by natural, production, and social factors. The multi-pond systems layout simulation results can be used to determine both the sub-watershed where ponds are needed to be set and the type of the set ponds, and the results show that the optimization layout could process the TP emission effectively and cost-efficiently, also the allocated pond system aligns with local production, lifestyle, and cultural practices. The developed method serves as a reference for utilizing agricultural multi-pond systems to manage nonpoint source pollution.</p>