It has been noted that high inflation has a significant threat to many countries developments and the persistence of inflation increases the cost of monetary policy in controlling inflation. Therefore, this paper proposes a unique method to estimate inflation persistence by using quantile autoregressive (QAR) model and to show how various shocks may affect the rate of inflation within different quantile levels. Based on the QAR model, the monthly inflation persistence in China and its dynamics over time are considered in the paper. The data illustrated here are the monthly year-on-year Chinese consumer price index from January 1987 to May 2019. The results indicate that the Chinese inflation series is globally stationary, but exhibits non-stationary behavior in about 44.85% of the observations. Again, it is evidence that the Chinese inflation rate has asymmetric characteristics at different quantile levels in its conditional distribution. Our research indicates that the QAR model outperforms the ordinary least square (OLS) method in terms of heterogeneity association of the inflation dynamics.