Objectives: Mild cognitive impairment (MCI) is known to have a high rate of progression to Alzheimer’s disease. Early detection and intervention of MCI are of great interest in psychiatric and socioeconomic aspects. There are various screening tools for MCI, but their sensitivity and specificity vary greatly. This study assessed the usefulness of virtual reality (VR) neurocognitive tests as an assessment tool for neurocognitive function deficit in MCI. Methods: Both VR neurocognitive tests and conventional neurocognitive tests, including MiniMental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), and the Seoul Neuropsychological Screening Battery (SNSB), were conducted, and 21 participants completed the tests. The test results of the MCI and normal groups were compared, and correlation coefficients between the VR neurocognitive tests and SNSB were examined. Results: The mean VR neurocognitive test total score of the MCI participants was significantly lower than that of normal participants (30.0±1.0 vs. 36.9±6.4; p<0.001). There were no significant differences in the SNSB, MMSE, and MoCA scores between the two groups. The VR neurocognitive total score correlated significantly with the MMSE, MoCA, and SNSB total scores (r=0.61, r=0.54, r=0.50, respectively; p<0.05). The scores of the subdomains of VR neurocognitive tests showed significant correlations with those of MMSE, MoCA, and subdomains of SNSB, with VR executive function and visuospatial function scores showing significant correlations with the SNSB executive function (r=0.46; p<0.05) and visuospatial function (r=0.60; p<0.01) scores, respectively. Conclusion This preliminary study suggests that the VR neurocognitive test can be a feasible and realistic tool for assessing the subtle but complex cognitive deficits in MCI, emphasizing spatial reasoning and executive functions.