International audience; The estimation of magnitudes of earthquakes that occurred before the instrumental period is a key issue in seismic hazard assessment. Such estimates are based on information provided by historical sources, which are translated into macroseismic intensity by means of intensity scales, which are then used, through different methodologies, to compute earthquake parameters. Within the AHEAD initiative we are beginning to analyze differences between methodologies regarding both magnitude and uncertainty estimates. For this initial exercise we relied on the information available in the SISFRANCE macroseismic database for a set of events located along the French-Italian border and compared the Boxer (Gasperini et al 2010) and QUake-MD (Provost and Scotti, 2017) methods.Although the two methods provide consistent results, we find a systematic magnitude bias between these two methods when using the epicentral information (location and epicentral intensity) provided by Boxer. However, within the limits of applicability of each methodology this bias disappears when using for the QUake-MD methodology, the epicentral location and intensity value of the SISFRANCE database. Furthermore, uncertainties estimates provided by the two methodologies overlap underlining the importance of quantifying uncertainties in parametric earthquake catalogues.In this presentation, we will discuss the preliminary results obtained when calibrating both methodologies with the same dataset. In this framework, requirements and challenges of calibration dataset selection will also be presented.