To improve the weakness of low localization accuracy and slow rate of convergence in the heuristic-based node localization algorithm, WSN node localization based on social spider optimization and opposition based learning is proposed. For reducing random search in the previous stage, a strategy of Bounding-box is firstly performed by determining a square area where the heuristic individual will be initialized. Then weighted center opposition based learning is introduced to original social spider optimization to find the best-estimated location of nodes and improve the global searching ability. Simulation results show that compared with the traditional algorithm, the proposed algorithm has faster convergence speed and higher node localization accuracy.