In order to improve the prediction accuracy of private car ownership, the data of private car ownership in China from 2005 to 2020 is analyzed by using time series analysis method, and a prediction model based on dynamic regression （ARIMAX）model is established. Lasso model and grey correlation analysis are used to get the main factors affecting private car ownership, and the main factors are introduced into autoregressive integrated moving average（ARIMA）model as regression terms. The ARIMAX model is established on the basis of the ARIMA model. Through the comparison of model prediction, it is found that ARIMAX model has better fitting effect, which is suitable for the prediction of private car ownership in China.