Knowledge and projection of farm numbers and the structure of their population is an important issue for agricultural economists and policy makers. Although Markov chain models have enjoyed decades of popularity in forecasting total farm numbers, they generally fail to provide a detailed insight of the farm population’s structure; to overcome this caveat we estimate a parametric distribution of the utilized agricultural area of French commercial farms. Our method provides detailed information on the structure of the population and accounts for the specificity of off-land farming. We also model the influence of variables such as the farm’s legal status, type of farming and farm holder’s age. The estimation leads to a relevant description of the entire population of professional farm. When compared with the 2005 Farm Structure Survey data, our simulations based on FADN data display a close match across a number of key variables.