OBJECTIVE: To examine the potential indirect effect of meal frequency on mortality via obesity indices. DESIGN: Prospective cohort study. SETTING: Korean Genome and Epidemiology Study. PARTICIPANTS: This cohort study involved 148 438 South Korean adults aged 40 years and older. RESULTS: Meal frequency at the baseline survey was assessed using a validated FFQ. Outcomes included all-cause mortality, cancer mortality and CVD mortality. Cox proportional hazards regression models were employed to examine the relationship between meal frequency and the risk of mortality. Mediation analyses were performed with changes in obesity indices (BMI and weight circumference (WC)) as mediators. In comparison to the three-time group, the once-per-day and four-times-per-day groups had a higher risk for all-cause mortality. The irregular frequency group had a higher risk for CVD mortality. Both once-per-day and four-times-per-day groups exhibited higher risks for cancer mortality. The effect of meal frequency on all-cause mortality was partially mediated by WC. For specific-cause mortality, similar mediation effects were found. CONCLUSIONS: The data suggests that three meals per day have a lower mortality and longer life expectancy compared with other meal frequencies. Increased waist circumference partially mediates this effect. These findings support the implementation of a strategy that addresses meal frequency and weight reduction together.