This paper presents a high-order stochastic collocation method for solving differential equations with random inputs. The method combines the strengths of Monte Carlo methods and stochastic Galerkin methods. It leverages the smoothness of the solution in the random space to achieve fast convergence, similar to stochastic Galerkin methods, but is easier to implement as it only requires solving deterministic problems at each collocation point, similar to Monte Carlo methods. The computational cost depends on the choice of collocation points, and several feasible constructions are presented. One construction, based on sparse grids, is weakly dependent on the dimensionality of the random space and is more suitable for high-dimensional problems. Numerical examples demonstrate the accuracy and efficiency of the stochastic collocation method. The method is applied to stochastic elliptic equations and time-dependent problems, showing its effectiveness in handling high-dimensional random inputs. The results indicate that the stochastic collocation method is more efficient than Monte Carlo methods and can provide accurate solutions with fewer collocation points compared to stochastic Galerkin methods.This paper presents a high-order stochastic collocation method for solving differential equations with random inputs. The method combines the strengths of Monte Carlo methods and stochastic Galerkin methods. It leverages the smoothness of the solution in the random space to achieve fast convergence, similar to stochastic Galerkin methods, but is easier to implement as it only requires solving deterministic problems at each collocation point, similar to Monte Carlo methods. The computational cost depends on the choice of collocation points, and several feasible constructions are presented. One construction, based on sparse grids, is weakly dependent on the dimensionality of the random space and is more suitable for high-dimensional problems. Numerical examples demonstrate the accuracy and efficiency of the stochastic collocation method. The method is applied to stochastic elliptic equations and time-dependent problems, showing its effectiveness in handling high-dimensional random inputs. The results indicate that the stochastic collocation method is more efficient than Monte Carlo methods and can provide accurate solutions with fewer collocation points compared to stochastic Galerkin methods.