Changes in the import price of goods will have an influence on the global economy and trade, corporate decision-making, and many other aspects. Therefore, the aim of this research is to forecast the future customs prices of US imports from China. The data comes from the Bureau of Economic Analysis. It can provide an important reference for enterprises, government, and academia to make better decisions. In this paper, due to the strong seasonality, the Seasonal Autoregressive Integrated Moving Average (SARIMA) model is a common technique that is generally used in analyzing and predicting seasonality and stationary time series data. Using the Sarima model to predict U.S. imports of goods from January 1986 to December 2016 and to compare the predicted data with original data. To identify the best model and find out the highest accuracy of the model, this paper utilizes the Grid Search method and calculates Akaike's Information Criterion (AIC), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and adapted MAPE (AMAPE) value as the criterion for selection. The result shows that in the next six months, the amount will still be seasonal, but the overall trend will show small fluctuations and will not change much.