In this paper, we discuss the necessity to take into account recovery processes in the field of vulnerability assessment, which is more and more widely used to help decision-making. Considering farm vulnerability to flooding, according to the literature, there is a need to better understand recovery after flooding to evaluate vulnerability. Nevertheless, quantitative methods, mainly developed to appraise damage completely neglect this. We present, in this paper, a modelling approach to assess vulnerability to flooding at farm level with quantitative indicators. The model takes into account processes which occur at farm scale until the farm recover to a state considered similar to its initial one. We first present the model and particularly focus on the difficulties related to modelling recovery processes. The description is illustrated by an application on a farm type in the Rhone River context. Then, we present damages born by several farmer profiles which illustrate the difference between taking or not into account recovery. These simulations show that, depending on the possibility to farmer to access external resources, damages endured can be much higher than direct damages usually evaluated when recovery is not taken into account. Finally, we discuss the interest and warnings to use this model and present remaining research challenges.