：In dense environments, it is challenging for multiple pursuers to capture an evader with an uncertain strategy. When the pursuers start from arbitrary initial positions rather than being evenly distributed around the evader, it becomes even more difficult for them to quickly approach and successfully coordinate to encircle the evader. To address this issue, we present a novel hierarchical framework to address the problem of multi-pursuer evader-approaching and encirclement in dense environments without any initial position constraint. At the high level, a novel task assignment method based on Dynamic Reallocate Review Consensus (DRRC) is proposed to reduce the intersection paths and further resolve allocation conflicts. At the low level, an effective distributed motion planning algorithm based on Follow the Optimal Gap Method (FOGM) is proposed, integrating adaptive safety factor and optimal reciprocal collision avoidance for multi-pursuer’ collision ‐free trajectories, guaranteeing that multi-pursuer can efficiently and safely encircle and capture the evader. Extensive simulation experiments in different hunting scenarios and two real-world experiments validate the applicability and practicability of our methods.