Objective: It has been reported that Mahjong, which facilitates episodic memory, attention, and executive functions, may prevent MCI and dementia; however, there are only a few studies that investigated the effect of Mahjong. The aims of this study were to compare the cognitive function at pre- and post-Mahjong Class, and to examine if there were various effects of Mahjong on different baseline cognitive function. Method: A total of 89 participants (72 ± 6.5 years old), who played Mahjong once a week for 20 weeks, were divided into two groups based on baseline Mini Mental State Examination (MMSE) scores (Group A [GA] ≦ 26, Group B [GB] ≧ 27). All participants completed MMSE, Frontal Assessment Battery at Bedside (FAB), Trail Making Test (TMT) Part A and B, and Geriatric Depression Scale (GDS) at pre- and post- Mahjong class. Results: GA showed significant differences in MMSE serial 7 (p < .01) and total score (p < .05), TMT-A (p < .05), and GDS (p = .05), while GB was significantly different on MMSE place orientation (p < .05) and TMT-A (p < .01) between pre- and post-Mahjong Class. There were significant group differences in MMSE serial 7 and total score (ps < .01) and FAB sensitivity to interference (p < .01) at post-Mahjong. Conclusion(s): Elevation of MMSE and TMT-A scores in both groups suggested potentially positive effects of Mahjong on brain activities. Additionally, improved mood found in GA might indicate that Mahjong not only affected cognitive functions but also provided opportunities to enjoy social life among the healthy elderly in Japan.