Advances in virtual reality technology have enabled the creation of virtual environments (VEs) with significantly high visual fidelity when compared to real environments (REs). In this study, we use a high-fidelity VE to examine two effects caused by alternating VE and RE experiences: "context-dependent forgetting" and "source-monitoring errors." The former effect is that memories learned in VEs are more easily recalled in VEs than in REs, whereas memories learned in REs are more easily recalled in REs than in VEs. The source-monitoring error is that memories learned in VEs are easily confused with those learned in REs, making discriminating the source of the memory difficult. We hypothesized that the visual fidelity of VEs is responsible for these effects and conducted an experiment using two types of VEs: a high-fidelity VE created using photogrammetry techniques and low-fidelity VE created with primitive shapes and materials. The results show that the high-fidelity VE significantly improved the sense of presence. However, the level of the visual fidelity of the VEs did not show any effect on context-dependent forgetting and source-monitoring errors. Notably, the null results of the context-dependent forgetting between the VE and RE were strongly supported by Bayesian analysis. Thus, we indicate that context-dependent forgetting does not necessarily occur, which will be helpful for VR-based education and training.