Over the past decade, research using virtual reality and serious game-based instruments for assessing spatial navigation and spatial memory in at-risk and AD populations has risen. We systematically reviewed the literature since 2012 to identify and evaluate the methodological quality and risk of bias in the analyses of the psychometric properties of VRSG-based instruments. The search was conducted primarily in July–December 2022 and updated in November 2023 in eight major databases. The quality of instrument development and study design were analyzed in all studies. Measurement properties were defined and analyzed according to COSMIN guidelines. A total of 1078 unique records were screened, and following selection criteria, thirty-seven studies were analyzed. From these studies, 30 instruments were identified. Construct and criterion validity were the most reported measurement properties, while structural validity and internal consistency evidence were the least reported. Nineteen studies were deemed very good in construct validity, whereas 11 studies reporting diagnostic accuracy were deemed very good in quality. Limitations regarding theoretical framework and research design requirements were found in most of the studies. VRSG-based instruments are valuable additions to the current diagnostic toolkit for AD. Further research is required to establish the psychometric performance and clinical utility of VRSG-based instruments, particularly the instrument development, content validity, and diagnostic accuracy for preclinical AD screening scenarios. This review provides a straightforward synthesis of the state of the art of VRSG-based instruments and suggests future directions for research.