The purpose of making efficient and flexible manufacturing systems is often related to the possibility to analyze the system considering at the same time a wide number of parameters and their interactions. Simulation models are proved to be useful to support and drive company management in improving the performances of production and logistic systems. However, to achieve the expected results, a detailed model of the production and logistic system is needed as well as a structured error analysis to guarantee results reliability. The aim of this paper is to give some practical guide lines in order to drive the error analysis for discrete event stochastic simulation model that are widely used to study production and logistic system.