Performance evaluation of communications systems is usually done by analytical techniques or simulation. The analytical approach is normally based on a probabilistic description of the system operations in terms of Markovian processes or queuing systems, but only is directly applicable for simple systems, or simplistic models of complex systems. Simulation is a powerful tool but requires enormous computations to yield accurate performance estimators. In this paper we present an alternative technique: Petri Nets, a graphical and mathematical modeling tool applicable to many systems. We used Stochastic Petri Nets to model and evaluate queuing systems and some protocols. Simple and accurate models were developed. The main definitions relating Petri nets in general, and to the particular classes of Petri nets are summarized in Section 1. The following sections contains queuing models (Section2 ), Stop and Wait and Aloha protocols (Section 3), DSP protocol (Section 4). Section 5 concludes the paper.