The aim of the present study is to improve the effectiveness of automotive diesel engine and aftertreatment calibration process through the critical evaluation of several methodologies to estimate the soot mass flow produced by diesel engines fueled by petroleum fuels and filtered by Diesel Particulate Filters (DPF). In particular, its focus has been the development of a reliable simulation method for the accurate prediction of the engine-out soot mass flow starting from Filter Smoke Number (FSN) measurements executed in steady state conditions, in order to predict the DPF loading considering different engine working conditions corresponding to NEDC and WLTP cycles.In order to achieve this goal, the study was split into two main parts:Correlation between ‘wet PM’ (measured by soot filter weighing) and the ‘dry soot’ (measured by the Micro Soot Sensor MSS). Test activities have been carried out taking into account different boundaries conditions such as calibration, driving cycle, sampling probes positions;Identification of a reliable and accurate method that allows estimating the ‘dry soot’ starting from the FSN measurements. Different equations available in literature have been investigated. According to the test results, it has been selected the one which allows to better correlate the simulation output with the MSS sensor measurement over different test cycles.Therefore, by linking the steady-state FSN measurements to MSS instantaneous readings, and in turn, the MSS cumulative reading to soot filter weighting, the study helped to contribute a simulation method which bridges the gap between the commonly available FSN readings and the reference technique used in the literature, the soot filter weighting, which for several reasons may not be easily available during the engine development process.Furthermore, the performed test activities lead the Authors to a few best practices for the soot measurement in both steady-state and transient conditions, including sampling probes installation. They allow to measure repeatable data during the test activities and avoid inconsistencies during the engine development.