<p>Virtual reality (VR) became the most used extended reality system in architecture, engineering, and construction domains. It offers advantages through its immersive and interactive interface. There is, however, a need to investigate both its strengths and weaknesses especially in relation to the claim that it is a close surrogate for real-world performances. This study reports the findings of a meta-analysis on distance estimation (DE) in VR. Distance estimation, essential for spatial perception, remains to be a problem in VR even with advanced head-mounted displays. The study questions whether VR can match the real-world performance in DE to highlight its shortcomings as well as its potentials. The meta-analysis includes 77 pieces of data from 29 studies and investigates whether DE accuracy has improved and how DE interacts with task type, task environment, and target range. The results indicate VR is still underperforming in DE, head-mounted display’s weight is the significant factor, and task type and task environment significantly interact with DE. We conclude VR needs to be specifically tailored regarding the needs of practitioners in architecture and engineering industry and that it is not yet a substitute for real-world performances.</p>