Logistics industry is closely related to many industries, and it plays an important role in promoting GDP growth and improving social and economic development. Using the past express business volume to forecast the future value is helpful to reflect the development trend of logistics industry. Based on the express business volume data of Jiangsu province from May 2013 to April 2021, a simple seasonal ARIMA（autoregressive integrated moving average）model and a multiplicative seasonal ARIMA model as well as two combined ARIMA models based on TOPSIS （technique for order preference by similarity to ideal solution）criterion are selected to fit and forecast the express business volume data from May 2021 to February 2022. The average relative errors of the three models' forecasts are found to be smaller, and the combined forecasting model outperform the other two seasonal ARIMA models.