The invention discloses a vehicle anti-collision early warning method based on an analytic hierarchy process and grey fuzziness. The problem that the safety distance is not accurate in the prior art is mainly solved. The method comprises the implementation steps that first, an analytic hierarchy process structure of a vehicle safety distance is established, and the weight vectors of all levels of influence factors are calculated; second, a grey fuzzy method is used for calculating the evaluation value of an index level influence factor; third, the accelerated speed of a vehicle and the reaction time of a driver are corrected according to the evaluation value of the index level influence factor; fourth, a kinematic formula of the vehicle is used for calculating the absolute safety distance and the relative safety distance of the vehicle; fifth, early warning levels are divided according to the absolute safety distance, the relative safety distance and the vehicle distance detected by a sensor. According to the vehicle anti-collision early warning method based on the analytic hierarchy process and the grey fuzziness, the influences of human factors, vehicle factors, road factors and environment factors on the vehicle safety distance are comprehensively considered, an anti-collision early warning of the vehicle is given out, and the accuracy of the vehicle safety distance and the safety of vehicle traveling are improved.