Previous studies that have attempted to model the participation decision of farmers in agri-environmental schemes have used a static framework where it was not possible to examine changes in the participation decision of farmers over time. This is rectified in this paper by utilising an 11-year panel that contains information on 300 farmers for each year. The structure of this dataset allows us to employ discrete time duration random effects panel data logit models to model the determinants of entering the Irish Rural Environment Protection Scheme (REPS). We introduce a dynamic element into a number of the models by using the random effects logit model estimator, with lagged dependent variables as additional explanatory variables. The results point to the fact that systems of farming that are more extensive and less environmentally degrading remain those most likely to participate in the REPS. In addition, the results highlight the fact that where no attempt is made to control for unobserved heterogeneity or path dependency the effects of the farm- and farmer-specific characteristics may be overestimated.