This paper deals with the empirical mean method, which is one of the most well-known methods of solving stochastic programming problems. The authors present their results obtained in recent years and discuss their application to estimation and identification problems.