While many studies have investigated depression risk factors, few attempts have been made to weight, and compare them. Therefore, we conducted a prospective comparison of a sample of subjects suffering from major depressive disorder and a group of healthy subjects. We compared classic risk factors with internal elements such as personality, family dynamics and health locus of control. We also looked for prognostic factors.Forty people with major depressive disorder (the MDD group) were randomly assigned to different treatment groups and followed for two years. In parallel, we followed a group of 21 healthy subjects (healthy group). At the beginning of the study, sociodemographic data were recorded and all subjects were asked to complete the Multidimensional Health Locus of Control (MHLC) scale, the NEO Five-Factor Inventory (NEO-FFI), and the Family Adaptation and Cohesion Scale (FACES III). During the study, subjects were regularly assessed using the Hamilton Depression Scale (HDS) and the Short Form Health Survey (SF-12).Of the 23 explanatory variables, 13 were statistically different (p≤0.05): age, gender, number of people living together, income, extravert personality and neuroticism, Internal HLC, Powerful others HLC, Adaptability of the current couple and the family of origin, and Cohesion of the ideal and nuclear family and family of origin. The accumulation of risk factors doubles the chances of suffering from MDD (odds ratio 1.905**). Independent of treatment, among the 13 variables, the first nine explain 34.1% of change in depression measured on the HDS scale (p<0.001).While the size of our sample limits the robustness of our results, our study suggests that some risk factors are also prognostic. The respective weights of these factors vary as a function of age group. Finally, some, such as health locus of control, family dynamics or extraversion, can be modified as an adjunct to pharmacological treatment.