OBJECTIVES: The objective of the study was to identify different GP's practicing profiles, in order to better understand the way French GP's react to governmental measures to contain cost or to increase quality of care. METHODS: We constructed a large database about of 4700 GP's from 2 French regions (Aquitaine and Burgundy) for the year 2000. Variables described physicians' volume and structure of medical activity (office consultations, home visits, medical procedures, prescriptions of pharmaceutical, lab tests and X-rays examinations), income level, personal characteristics (age, gender, speciality, localisation, contractual status with health insurance), practice characteristics (size, structure by age, percentage of hospitalised patients, etc.) and socio-economic characteristics of the geographical environment (physicians density, distance to the nearest hospital, unemployment rate, urban or rural town). The database was analysed though a variety of statistical methods including cluster analysis and econometric tests. RESULTS: Four different homogeneous groups were identified, each of them relating physicians’ level of activity to their socio-economic status. In group 1 for instance, GP's have a high level of activity (2340 patients and 9300 consultations in average), 46% of them practice in rural areas and only 4% are women. Conversely, physicians in group 4 have a low level of activity (1050 patients, 2400 consultations in average), 75% are located in urban areas and 44% are women. Econometric tests allow discriminating the main determinants of multidimensional medical activity between rural and urban GP's. CONCLUSIONS: The way GP's exert medicine is not uniform. The level and the type of medical activity greatly vary among physicians. Individual factors as well as characteristics of the socio-economic environment greatly explain these differences. An immediate consequence is that any cost-containment measure which applies uniformly to all GP's, such as regulating fees, necessarily results in different outcomes according to the type of physicians’ category.