In view of the shortcomings of the traditional recommendation algorithm in the loan portfolio of P2P platform, a loan portfolio recommendation model based on multi-objective evolutionary algorithm is proposed. To balance the accuracy and the rate of return of the recommended loan portfolio, the objective function is established to predict the rating and consistency of the recommended loan portfolio. The objective function is solved based on the improved multi-objective evolutionary recommendation algorithm. To solve the problem of high dimension of the decision variable space, the dimension reduction of the decision space and the improved initialization strategy are proposed to accelerate the convergence speed of the population and improve the search efficiency of the algorithm. The proposed model is compared with collaborative filtering (CF), neural collaborative filtering (NCF), probabilistic spreading (ProbS), particle swarm optimization (PSO), dynamic programming algorithm (DPA), MOEA-ProbS and other models, and the results show that the performance of the proposed model is improved, the average accuracy is 0.104 7, the average profit coefficient is 0.154 2, and the average risk coefficient is 0.002 3.