The accurate prediction of critical heat flux (CHF) is of great significance to the safe operation of the reactor. At present, most of the traditional methods for predicting CHF are based on relational fitting in mathematical form, which are not accurate enough and may can not taking full advantage of experimental data. Therefore, a prediction model based on random forest algorithm (RF) is proposed, the model is based on CHF 2006 look-up table, and the accuracy and smoothness of the model have been verified when predicting CHF. In this paper, the variation trend of CHF predicted value and actual value with pressure, mass flow rate and thermodynamic quality is given. It has been verified that the error of this method is acceptable, the predicted value is in good agreement with the experimental value, and conforms to the corresponding physical laws, which can provide a reference for the development of high-precision CHF prediction models or CHF-related experiments.