This article presents the PROMESS projection model, which can be used to assess certain consequences of pension reforms by performing a quantitative and aggregate analysis of the overall population. The model draws on recent micro-econometric data concerning retirement behaviour and on a very large wealth of available statistical data related to the composition of the population (by gender and generation, as well as by number of validated quarters and average wage quartiles). The PROMESS model can be used to quantify the impact of the 1993, 2003 and 2010 French pension reforms on the employment rates of elderly persons, as well as on retirement ages (differentiated by population subgroup). These reforms appear to have a significant impact on activity after the age of 60, but also before. Indeed, they induce a 4-percent long-term rise in the employment rate of 55-59 year olds due to the impact of the 1993 and 2003 reforms, and a 5 % rise due to the 2010 reform. On the other hand, they seem to induce overall redistributive effects in terms of exits from working life (without, however, distinguishing between the effects of each reform) : the average retirement and pension windingup postponement tends to increase proportionately to wage level.