Stress is known to elicit contrasting patterns of plasticity in the amygdala and hippocampus. While chronic stress leads to neuronal atrophy in the rodent hippocampus, it has the opposite effect in the basolateral amygdala (BLA). Further, even a single episode of acute stress is known to elicit delayed effects in the amygdala. For example, 2 h of immobilisation stress has been shown to cause a delayed increase in dendritic spine density on BLA principal neurons 10 days later in young rats. This is paralleled by higher anxiety-like behaviour at the same delayed time point. This temporal build-up of morphological and behavioural effects 10 days later, in turn, provides a stress-free time window of intervention after exposure to acute stress. Here, we explore this possibility by specifically testing the efficacy of an anxiolytic drug in reversing the delayed effects of acute immobilisation stress. Oral gavage of diazepam 1 h immobilisation stress prevented the increase in anxiety-like behaviour on the elevated plus-maze 10 days later. The same post-stress intervention also prevented delayed spinogenesis in the BLA 10 days after acute stress. Surprisingly, gavage of only the vehicle also had a protective effect on both the behavioural and synaptic effects of stress 10 days later. Vehicle gavage was found to trigger a significant rise in corticosterone levels that was comparable to that elicited by acute stress. This suggests that a surge in corticosterone levels, caused by the vehicle gavage 1 h after acute stress, was capable of reversing the delayed enhancing effects of stress on anxiety-like behaviour and BLA synaptic connectivity. These findings are consistent with clinical reports on the protective effects of glucocorticoids against the development of symptoms of post-traumatic stress disorder. Taken together, these results reveal strategies, targeted 1 h after stress, which can prevent the delayed effects of a brief exposure to a severe physical stressor.