Traditionally, the Common Agricultural Policy has used minimum intervention prices and coupled payments to support farmers in Europe.  These highly trade-distorting instruments have received much criticism.  Following the Uruguay Round, a major reform process started in the EU.  In 2003, the EU decoupled farm support by introducing Single Farm Payments.  However, economic theory has shown that such payments may lead to increases in farm output because of the existence of market imperfections or risk.  This thesis explores the issue of farm support for the case of crops farm households in France.  It develops and applies a stochastic dynamic farm household model for a representative crops farm household.  The farm household faces occasionally binding debt constraints and investment adjustment costs.  The model is specified and parameterized to a sample of French crops farms using Farm Accountancy Data Network (FADN) data observed between 1996 and 2003.  Simulations with various policy scenarios show the extent to which farm support and especially decoupled payments affect the farm household’s decisions.  First, the thesis examines the impact of CAP reforms on farm behaviour.  Then, the thesis compares the impacts of different farm payments, i.e. Single Farm Payments, farm payments coupled to production and Counter-cyclical Payments, to explore the link between investment and farm payment.  Further, the thesis highlights that the degree of decoupling of farm payments depends on the farm household’s access to credit.  Although Single Farm Payments may potentially affect output decisions, given the relatively low impacts found in the thesis, it is reasonable to consider that the Single Farm Payment is effectively decoupled in this French FADN sample of arable crops farms.  The introduction of decoupled direct payments in the EU appears to be an efficient step towards reducing trade distortions.