The present paper portrays the mapping for stochasticity in low-velocity impact responses of functionally graded material (FGM) plates by employing multivariate adaptive regression splines (MARS) surrogate model in conjunction with finite element (FE) approach. The unavoidable stochastic variabilities (caused due to numerous errors involved in manufacturing processes) in material properties of FGM plates are considered in order to map the effect of elemental variabilities on global response of the structure. The material properties of FGM plates are considered to follow the rule of mixture in conjunction to power law. The Newmark’s time integration scheme and modified Hertzian contact law are employed to solve the time-dependent equation. The present FE formulation is based on an eight noded isoparametric element in which each element has five degrees of freedom. The effects of variability in temperature and power-law exponent on stochastic low-velocity impact responses are also portrayed. The maximum contact force, plate and impactor displacement are considered as the response parameters. The present MARS model is coupled with the finite element to achieve the higher efficiency with adequate accuracy as compared to the FE-based full-scale Monte Carlo simulation. The statistical results illustrate that the stochasticity in material properties significantly influences the low-velocity impact responses of FGM plates.