This article presents the result of an experiment on rote memorization, performed in two virtual reality learning environments, 3D (VR 3D) virtual reality and virtual reality based on 360° images (VR 360). 201 participants took this test aiming at memorizing the capitals of unusual countries. Feedback was performed immediately after the test, by filling in an online questionnaire on a desktop. The obtained results show that the VR 3D format (29.6% of capitals recalled) is less efficient than the VR 360 (30.8%) for rote learning purposes. Moreover, the average memorization result for both formats (3D and 360) remains lower than the result obtained by another experiment which used paper format to perform the same task. Scores achieved using VR 360 (30.8%) are lower than when using paper (36.8%), that is 6.8 pts less for VR 360. Therefore, the format characteristics tend to impact memorization, which for VR 360 consisted in displaying pairs of capital-country in situ, on 360° images depicting the given capital. Furthermore, the experiment tends to reveal that in immersive universes, the subject's age and involvement in an academic environment (pupils, students, professors) has no effect on the results, whereas the time spent on a task produces contrasted results.