Purpose: The purpose of this study was to investigate the relationships between regular exercise, cognitive function, and depression in Korean elderly people, and to explore whether age group affects the relationships between these variables. Methods: This study utilized data from the Korean Longitudinal Study of Aging (KLoSA) from waves 1 to 6. The sample in the current study was 4244 participants, aged 65 and older who were in wave 6 (in 2016). Welch’s test were conducted to identify the relationships between regular exercise, cognitive function, and depression. Results: First, participants undertaking exercise had better cognitive function than non-exercising participants. Second, old-old elderly people who had started exercising in the last 2 years after at least 8 years of having an inactive lifestyle had better cognitive function than those who have not exercised for more than 10 years. Third, participants who exercised had lower levels of depression than those who did not exercise. Fourth, old-old elderly people who started exercising in the last 2 years after at least 8 years of a sedentary lifestyle tended to have lower levels of depression than those who had not exercised for more than 10 years. Conclusion: The current study extends knowledge on the importance of regular exercise on cognitive function and depression in elderly people. Furthermore, this study suggests that the positive effects of exercise on cognitive function and depression may be greater for the old-old compared to the young-old age groups.