Aiming at the shortcomings of current prediction model of Portuguese performance, such as low accuracy and slow speed, a performance prediction model of Portuguese on the basis of the particle swarm optimization neural network is proposed. First of all, a large number of Portuguese performance data are collected and preprocessed, then the neural network is used to study the teach- ing samples of Portuguese performance, and particle swarm optimization algorithm is used to select the threshold value, weight value and other parameters of the neural network, so as to establish the optimal prediction model of Portuguese performance, in the end, the effectiveness and superiority of the model are compared and tested with the performance data of Portuguese. The results show that this model can improve the accuracy and reliability of the prediction results, and provide valuable in- formation for Portuguese teaching.