Network researchers have dedicated a notable part of their efforts to the area of modeling traffic and in particular to the definition of rigorous statistical models. We feel that there is a strong demand for traffic generators capable to reproduce realistic application traffic according to theoretical models. Indeed, nowadays the trend is changing and network researchers are melting together networking and stochastic knowledge in order to fill in the actual gap. This work presents an integrated set of tools we called Internet Traffic Generator (ITG), which allows to reproduce TCP and UDP traffic and to accurately replicaie appropriate stochastic processes for both IDT (Inter Departure Time) and PS (Packet Size) random processes. We believe that ITG shows interesting properties when compared to other traffic generators.