Abstract Active nuclear safeguards measurements that rely on the time correlation between fast neutrons and gamma rays from the same fission are becoming a useful technique. In previous works we have shown the feasibility of this method, in conjunction with the use of the well-known MCNP simulation code and the use of artificial neural networks, to estimate the mass and enrichment of fissile samples enclosed in special, sealed containers. In a more recent works, we pointed out some features of MCNP that represent a drawback in the simulation of correlation measurements. In fact, MCNP is not intended for obtaining second order moments. Therefore, to achieve an agreement between the simulations and the experiment, we resorted to the use of effective physical parameters (for example detection threshold and light output). Recently, by suitably modifying MCNP, we developed the MCNP-PoliMi code, which attempts to simulate the physics of each interacton more realistically. In this paper, we present the results of the calibration of a plastic scintillator using a Cf-252 source, traditionally used in the above-mentioned nuclear safeguards experiments. We show that the physical parameters found by calibration and used in the MCNP-PoliMi simulations are reasonable for this type of detector, and that the correlation functions thereby obtained are in good agreement with the experiment.