This work is composed of four parts : the first one is about the new geography and its active participation in the field of health, the second presents the Nord – Pas-de-Calais region, the national and regional programs of health care planning and develops some methods used in our modeling : a partitioning by K-Means and Relative Neighbourhood, Gabriel and Delaunay triangulation connection networks. The third part presents materials, methods, results and discussion of our modeling process in four phases of the spatial utilization of health services. The fourth part is a general discussion on the geographical prospective of the gravitation models and on the geography of health's booming in France. Our model assesses the health care demand and supply in our Nord – Pas-de-Calais region. It calculates the theoretical hospital catchment areas by specialty in hematology and traumato-orthopedy. It will be automated and generalized for all specialties and geographic areas. The first phase observes the hospital catchment areas with quantitative and cartographic tools. The second is an elaboration of our spatial mathematical model regarding the gravitation Newton law. The originality of it is our weighting extension of the Reilly equation based on the demographic population of proximity catchment area and calculated from the observed and theoretical hospital attractions. The third phase is a validation of our model by comparison between the observed and the theoretical models with the help of observed maps, models drawings and quantitative methods. The last predicts and simulates “the creation or the closing” of hospital services by specialty. Through results this model appears to be robust, reliable and predictive in spite of its simplicity (two parameters are used : numbers of beds and the demographic population). Consequently, our model, along with a better knowing of the spatial utilization of health services, can be a new tool for developments in health care districts.