Despite numerous works about the structure of capital, we still ignore on which bases enterprises make their financing choices (Myers, 1993). The case of France is in this matter of particular interest, because according to various works, French enterprises are alleged to have reduced their debts over the last decade (Laboureix & Laurin, 1994; Boissonnade & Tournier, 1996). First, we show that this is not true: over the 1988-1996 period, some enterprises have got into serious debts, while others reduced their debts. Therefore, we are trying to modelize and to test short-term financial behaviors, by proposing a dynamic analysis. This constitutes an original contribution, since to this day the French situation has only been the object of a few cross section studies (Dubois, 1985; Biais, Hillion & Malécot, 1995; Rajan & Zingales, 1995). We argue that this strategy is constructed around the notion of target-ratio, a notion only recently put forward (Fama & French, 1997; Opler & Titman, 1996; Shyam-Sunder & Myers, 1999), and we analyze the relationship between changes in indebtedness and a vector of explanatory variables, through univariate methods, as well as parametric and non-parametric multivariate methods. Our results clearly show taht target-ratio considerations guide the strategies of French enterprises. All tests confirm that the ECARTS variable, which represents the difference between the real indebtedness ratio at the beginning of a period and the target-ratio, is positively associated to the decrease in debts. This remains true when the other classical explanatory parameters of the evolution in debts are0501ntained constant. Profitability and size, which are both inversely related to the increase in debts, appear as the other two important determinants of the evolution of indebtedness.