Approximation by Poisson distributions or processes is of great interest in many applications including networks, queueing theory and even neuroscience, as the Poisson distribution arises as a natural limit distribution in many situations involving independent “rare” events. In this talk, we propose to give an overview of classical methods of Poisson approximation of sums of Bernoulli random variables. We will first focus on Le Cam’s result for the case of independent random variables and give a proof based on the coupling method. We will then focus on the case when the variables are not necessarily independent and present the classical Chen-Stein method.