Abstract: To solve the problem of the congestion of urban traffic network, a fuzzy evaluation model of urban road traffic congestion was developed based on the fuzzy analytic hierarchy process and neural network theory. Firstly, taking AHP as the criterion, a road traffic congestion evaluation index system was established with three levels, including microcosmic, meso and macro layer. Then a fuzzy consistent judgment matrix was used to define the evaluation factor weights of road traffic congestion, a fuzzy neural network evaluation model was built based on BP neural network to judge the road traffic congestion situation, the threshold intervals of the road traffic congestion evaluation were determined as five grades, including moderate congestion, severe congestion, mild congestion, relatively smooth and smooth. The traffic congestion data of 20 traffic districts in Shanghai in 2016 were taken as example to implement the method. The results show the feasibility and effectiveness of the method.