Stress before retention testing impairs memory, whereas memory performance is enhanced when the learning context is reinstated at retrieval. In the present study, we examined whether the negative impact of stress before memory retrieval can be attenuated when memory is tested in the same environmental context as that in which learning took place. Subjects learned a 2-D object location task in a room scented with vanilla. Twenty-four hours later, they were exposed to stress or a control condition before memory for the object location task was assessed in a cued-recall test, either in the learning context or in a different context (unfamiliar room without the odor). Stress impaired memory when assessed in the unfamiliar context, but not when assessed in the learning context. These results suggest that the detrimental effects of stress on memory retrieval can be abolished when a distinct learning context is reinstated at test.