Background: Several studies support the effectiveness and tolerability of virtual reality (VR) interventions in the psychiatric field. This study aimed to examine changes of cognitive performance in VR-assisted mental health promotion programs and to investigate the difference in performance according to clinical characteristics. Methods: Thirty subjects aged >55 years participated in the study. The clinical characteristics of depression, anxiety, perceived stress, quality of life, and cognition were assessed. Cognitive performance in VR-assisted mental health promotion programs was compared between the clusters classified by clinical characteristics. Results: Cluster analysis classified the subjects into three groups. In Cluster 1, the Module 3 training score was significantly different before and after VR performance. In Cluster 2, significant differences were observed in the Module 1 training score, the Module 2 training score, the Module 2 defense failure score, and the Module 3 training score. In Cluster 3, a significant difference was observed in the Module 3 training score. Conclusion: The results of this study suggest that VR performance might differ according to clinical characteristics. A cognitive training strategy using VR has to be differentially established depending on the characteristics of the community population.