Weapon target assignment (WTA) is a representative complicated combinatorial problem widely investigated and used. Evolutionary Algorithms (EAs) are often adopted to resolve combinatorial optimization problems including WTA. The crow search algorithm (CSA) is an excellent EA which simulates the crows’ behavior of hiding and thieving food. To resolve WTA efficiently, an enhanced CSA is proposed. It has two developed methods, i.e., neighborhood search strategy and portion reinitialization method. Neighborhood search strategy uses four variable swapping approaches to enhance algorithm’s global and local optimization abilities simultaneously. Portion reinitialization method is introduced to reinitialize part of non-promising individuals when the algorithm is trapped into local optima to improve global optimization performance. Twelve classical WTA benchmark instances and six algorithms are selected to evaluate the proposed algorithm, and the results demonstrate that our method has significantly superior ability and satisfying time cost.