Introduction To memorize a path and then be able to retrace its steps in a new environment, one of the strategies consists in memorizing a succession of landmarks associated with a given direction. It has been shown that Alzheimer's disease (AD) leads to a deficit in route memory, from the early stage of the disease [1-2]. However, health care facilities in which the studies are generally conducted have been described as poor in terms of land-marks and may thus increase disorientation. The purpose of this comparative study was to assess the structuring effect of the elements of a real environment, rich in landmarks, on route learning (forward and return trips), in a population of AD and healthy control (HC) subjects. 2 Methodology Thirty subjects at mild to moderate stage of the disease and 30 matched HC underwent route learning: after 1 learning route (forward trip), they were asked to reproduce it in the same direction 2 times, then a third time in the opposite direction (return trip). The task was carried out in the healing garden "art, memory and life" of the CHRU of Nancy, designed to contribute among others, to mitigate spatial disorientation [3]. 3 Results As expected, subjects with AD committed significantly more errors than the HC, re-gardless of the trip. However, contrary to the literature data, the AD subjects did not make more mistakes in the return trip than in the forward trip, which is a more complex task [4-5]. In addition, the analysis of the location of the errors along the route showed that they were not located in the same areas of the route in forward and return trips. According to the qualitative analyzes, the salience of the elements of the environment but also their "perceived affordance" [6] (their possibilities of action) have influenced the directional choices and this in a different way according to the trip (forward-return) and the group (AD-HC). 4 Discussion After several trials, the subjects with Alzheimer have thus acquired some knowledge of the "route" type. The nature of the environment, rich in landmarks, seemed to favor this acquisition. To reproduce a route requires, at least, memorizing a direction associated with a given place / landmark. This association was favored by the "perceived affordance" of the elements which depended on: 1) the point of view of the subject and thus, varied according to the direction of the trip: backward/return; 2) cognitive control abilities, and so, varied by group: HC/AD. This study shows that the elements of the environment have a structuring effect on learning performance that is not absolute but relative to these 2 factors: direction of the trip and the cognitive control abilities of the subjects.