: In order to determine the risk level of collapse in the highway tunnel construction process, four primary indicators such as engineering geology and 12 secondary indicators such as surrounding rock grade are selected to establish an entropy-improved gray correlation highway tunnel collapse risk assessment model. The entropy weight method is used to calculate the weight of the evaluation index; the improved gray correlation method is used to determine the correlation degree between each section of the tunnel and the risk level of the tunnel collapse, and the final risk level is determined by combining the obtained weights, and the result is compared with the engineering example. The results show that among the 12 risk factors of highway tunnel collapse, annual average precipitation, tunnel span, surrounding rock grade and tunnel depth have great influence on tunnel collapse; the four-segment tunnel is evaluated by this model, and the evaluation results are both actual and actual. Match. The evaluation model provides a scientific and effective basis for the determination and prevention of the collapse risk level of tunnel construction under construction.