Abstract: The future performance evaluation of landslide lake risk disposal could provide a scientific basis to improve risk disposal of landslide lake. The intelligent evaluation system based on the improved Compensation Fuzzy Neural Network (CFNN) was introduced to the future performance evaluation of landslide lake risk disposal, to solve fuzzy and non-linear evaluation problem of stability, time, cost and ecological index. Through the simulation experiments, the future performance evaluation on improved CFNN which had high convergence speed, fault tolerance and adaptive ability, was an effective evaluation method of landslide lake risk disposal.