As in many complex behavioral responses, inter-individual variability can be observed in the responses to a chronic mild stress. While some subjects exhibit more resilient behaviours, others appear more susceptible to stress. This study hypothesizes that this variability relies on the individual appraisal of the stressful event. To study this assumption, mice were first subjected to a conditioned task occurring in a circular arena. In this task, a mild air-puff (i.e. stressor) in a given quadrant of the arena was coupled with the presence or the absence of a light in the same quadrant. Half of mice were then submitted to a 15-day subchronic stress consisting in various environmental and social mild stressors randomly applied two or three times a day. At the end of this procedure, the occurrence of depressive-like behaviours in stressed mice was assessed using measures of the stress regime (i.e. physical state, choice test, grooming test). The physical state assessed the physical appearance of mice. The grooming test consisted in measuring the time spent in grooming after mice were sprayed upon with a viscous solution. The choice test consisted in measuring the time spent in an uncomfortable place (i.e. whose floor was covered with damp sawdust) versus a more comfortable one (i.e. with dry sawdust) to evaluate the reactivity to a negative stimulus previously encountered during the subchronic stress. Multiple regression analyses revealed a relationship between attention toward salient stressful stimuli in the conditioned task and susceptibility to the subchronic stress procedure. These results are discussed regarding their relevance for the understanding of aetiologies of depressive illnesses.