Based on the conditions of water resources endowment, the efficiency principle, and the principle of respecting actual conditions, an index system for water pollutant load allocation and a water pollutant allocation projection pursuit(PP)model were established. It is difficult to determine the optimal projection direction using the PP model. In order to solve this problem, the sine cosine algorithm(SCA)was used to search for the optimal projection direction of the PP model. An SCA-PP model was built for the water pollutant load allocation for eight counties(cities)in Wenshan Zhuang and Miao Autonomous Prefecture of Yunnan Province. The SCA algorithm was verified using six typical test functions. The simulation results were compared with those of the ant colony optimization(ACO)algorithm, simulated annealing(SA)algorithm, cultural algorithm(CA), Cuckoo search(CS)algorithm, and artificial bee colony(ABC)algorithm. The results are as follows: (1)The SCA algorithm performs better than the ACO, SA, CA, CS, and ABC algorithms. It is simple and has a fast convergence speed, high precision of optimization, high capability of global optimization, and strong convergence stability and reliability. (2)The simulations results of the SCA-PP model for water pollutant load control meet the economic and social development and water pollutant reduction requirements in the study area. The model has certain operability and effectiveness, providing a new method of water pollutant allocation.