The Error of a Function Approximation Based on Random Selected Points Abstract: The advantages of the Monte-Carlo quadrature for solving least-square approximation problems are discussed, an approximation algorithm and a probabilistic error estimate are given.