Abstract: Abstract：[Objectives] Focusing on underactuated marine vehicles swarm tracking control problem with UMVs unmodeled dynamics, external environment disturbance and complex swarm reference signals, UMVs swarm motion decision and finite-time swarm tracking control scheme are studied. [Methods] Firstly, using the Lyapunov function and artificial potential function, combined with the position information of the swarm virtual reference vehicle (SVRV), a swarm motion decision strategy with velocity guidance is constructed. Secondly, a finite-time lumped uncertainty observer (FLUO) is designed to compensate for the unknown information contained in the velocity error equation. Furthermore, a non-singular terminal sliding mode (NTSM) swarn tracking control method based on FLUO is adopted to design the control law. [Results] Based on the Lyapunov stability theory and Matlab/Simulink simulation experiments, it is proved that the system is stable. [Conclusions] Based on the proposed swarm decision strategy and FLUO-NTSM control method, UMVs can keep swarm and achieve accurate path tracking.