The invention discloses a method and a system for assessing values of second-hand cars according to big data. The method includes S11, acquiring model information, input by users, of to-be-accessed second-hand cars; S12, acquiring corresponding value assessment functions from a function database of a value assessment model, and after acquiring all corresponding function parameters from a parameterdatabase of the value assessment model, substituting the vehicle information into the value assessment functions to calculate market assessed prices of the to-be-accessed second-hand cars. The modelinformation includes models, license application dates, mileages, vehicle conditions, trading locations and/or colors; the value assessment model is a multi-element nonlinear calculation model acquired based on big data training, and comprises the function database and the parameter database. The method and the system have the advantages of high value assessment accuracy, high stability, capability of assessing the prices of the second-hand cars accurately, and wide application to the value assessment field of the second-hand cars.