In order to predict perforation velocity of SCS (steel-concrete-steel) slab against missile impact and obtain influence sequence of structural factors of SCS slab on the perforation velocity, a dimensionless equation of the perforation velocity was established and a prediction model was obtained based on dimensional analysis and artificial neural network. Orthogonal experiment design was used to determine finite element calculation, and the influence degree of 7 factors of SCS slab on perforation velocity was quantitatively evaluated by variance analysis. The deviation between the predicted perforation velocity and the actual value is less than 12%, and the quantitative evaluation results of variance analysis show that the thickness of steel plate has a largest effect, followed by the distance of tie bar and the thickness of concrete, the yield strength of steel plate, the yield strength of tie bar and the diameter of tie bar have a smaller effect, and the concrete compressive strength has a smallest effect. The established model solves the prediction problem of perforation velocity from a new perspective with a good prediction effect, which can effectively evaluate the ability of SCS slab against perforation failure of missile impact, and the order of the factors is beneficial to the optimal design of SCS slab resisting perforation from missile impact.