The invention discloses a method for intelligently predicting price of secondhand cars, and belongs to the technical field of artificial intelligence. The method comprises the following steps: according to influence factors of the price of secondhand cars, selecting attributes required by regression of the price of the secondhand cars, and quantifying the attributes; screening out attribute characteristic vectors and past transaction data, and training a decision regression tree model; setting an update time interval and newly added transaction data indexes, and updating models regularly; selecting the car according to attributes including brands, car series, configuration, gearbox category, color, car type, displacement, purchase time, road mileage, trade frequency, newness rate, the condition whether the car is subjected to serious collision, the condition whether the car is subjected to water immersing for a long time and the condition whether the car is subjected to fire. Accordingto the method disclosed by the invention, key factors influencing the price of the secondhand cars are selected, and the decision regression tree model is trained by the transaction data, so that thepurpose of intelligently predicting the price of the secondhand cars is achieved, the price evaluation of the secondhand car is more objective and fairer, and the fair dealing of the secondhand car is effectively guaranteed.