Site effects represent a fundamental aspect in the definition of the seismic action. For ordinary design applications, several building codes allow the estimate of the seismic ground response through simplified approaches. The aim of this study is the assessment of the simplified method proposed in the Final Draft of revision of the Eurocode 8. For this purpose, equivalent linear analyses have been performed on a collection of 1-D ground models, generated from a database of real soil profiles through a stochastic procedure. Firstly, the study evaluated the inter-class dispersion of the new site categorization system, observing a reduction of the variability. Then, it focused on the reliability of the amplification factors. These values provide a good prediction of the seismic action, when referring to a wide range of vibration periods. However, their adoption gives an underestimation at short periods, whereas the estimate is on the safe side at long periods.