Abstract: Objectives To address the requirements of formation description and maneuverability for multiple unmanned surface vehicle (USV) systems in irregular working environments, a distributed cooperative maneuvering target tracking control method is proposed based on affine transformation. Methods First, a novel output redefinition-based dynamic transformation method is proposed to resolve the lack of relative degree in underactuated USV systems, thereby simplifying control design. Subsequently, a distributed target velocity observer is designed to estimate the unknown velocity of moving targets. Leveraging the estimated target velocity, a distributed target tracking control strategy is developed by combining affine transformation theory and backstepping control technology to achieve maneuverable tracking of moving targets. To address unknown model parameters and external disturbances, an adaptive law combining the minimal learning parameter (MLP) technique and radial basis function neural network (RBFNN) is designed for online estimation. Results Simulation results demonstrate that the proposed control method achieves effective tracking of moving targets in irregular working environments. When targets traverse narrow passages, the multiple USV system flexibly adjusts its formation to avoid obstacles, exhibiting robust control performance and flexibility. Conclusions The proposed affine transformation-based control method enhances the flexibility and safety of multiple USV systems in moving target tracking tasks within irregular working environments. This work provides a significant advancement in distributed cooperative control strategies for multiple USV systems, with practical implications for maritime surveillance and rescue operations.