The modelling of the performance of asphalt pavement is of great importance in the scientific development of maintenance strategies. The Pavement Condition Index (PCI) and the Pavement Comprehensive Evaluation Index (PCI) are both common pavement performance evaluation indices. The traditional prediction method is difficult to accurately predict the performance of asphalt pavement in Xinjiang, due to the region’s unique climate and traffic conditions. This paper proposes an asphalt pavement performance decay modelling method based on Holt’s exponential smoothing prediction method, with the PCI and the PQI as the study objects. The Holt double exponential smoothing model is utilised for the purpose of making predictions. The findings indicate that the Holt model exhibits minimal error in short-term prediction, effectively captures the decay trend, and provides data support for pavement preventive maintenance. The study further validates the efficacy of time series modelling in pavement management.