Abstract: Seismic liquefaction of sand is a dynamic geological phenomenon caused by the joint action of multiple influencing factors, and it is difficult to accurately distinguish the seismic liquefaction state of sand by conventional models. In this paper, the principal component analysis was carried out on the selected nine influencing factors of sand seismic liquefaction, and four principal components were extracted. At the same time, the support vector machine was introduced to establish the prediction model of sand seismic liquefaction. Combined with an engineering example, the prediction results were compared with the prediction results of optimized support vector machine model without principal component extraction. The results showed that the prediction model of sand seismic liquefaction based on principal component analysis and optimized support vector machine had higher accuracy, and could provide effective support for earthquake disaster prevention and control work.