The present work is part of the 2015 ART-IT project (Early Earthquake Alert in Italy). Its main purpose is the estimation of the PRESTo system [PRobabilistic and Evolutionary early warning SysTem, Iannaccone et al., 2010; Satriano et al., 2011] performance through an ad hoc calibration of the software configuration parameters using only the accelerometric networks installed in an area of Central Italy, selected as test site. The correct operation of an early warning system could be used for a more correct and effective management of a seismic emergency from the first seconds after the occurrence of a strong earthquake, allowing to adopt sudden actions to reduce the exposure and, consequently, the seismic risk. To evaluate the performance of the software, it is necessary to carry out an a priori calibration phase of the configuration parameters in order to guarantee the best performance in terms of seismic event detection and reliability in the estimation of hypocentral parameters. The tests were carried out by means of a series of recorded waveforms simulations considering a selected catalogue of earthquakes with a magnitude range between 3.9 and 5.4, occurred in the Amatrice-Norcia area, with a magnitude range between 3.9 and 5.4, occurred in the Amatrice-Norcia area, with particular focus on the seismic sequence of January 2017.