Background: The triangle completion test has been used to assess egocentric wayfinding for decades, yet there is little information on its reliability. We developed a virtual reality (VR) based test and investigated whether either test of spatial navigation was reliable.

Objective: To examine test-retest reliability of the real-world and VR triangle completion tests. A secondary objective was to examine the usability of the VR based test.

Materials and methods: Thirty healthy adults aged 18–45 years were recruited to this block randomized study. Participants completed two sessions of triangle completion tests in the real-world and VR on the same day with a break between sessions.

Results: In both test versions distance from the endpoint and angle of deviation showed poor test-retest reliability (r < 0.5). Distance traveled had moderate reliability in both the real-world and VR tests (r = 0.55 95% CI [0.23, 0.76]; r = 0.66 95% CI [0.4, 0.83, respectively]). The VR triangle test showed poor correlation with the real-world test.

Conclusion: The triangle completion test has poor test-retest reliability and demonstrates poor concurrent validity between the real-world and VR. Nevertheless, it was feasible to translate a real-world test of spatial navigation into VR. VR provides opportunities for development of clinically relevant spatial navigation tests in the future.