BACKGROUND: Depression is a major health problem among older adults, for which previous studies have suggested tooth loss as a risk factor. This study examined the mediating effect of oral function and orofacial appearance in the relationship between tooth loss and depression. METHODS: This three-year follow-up longitudinal study was based on a self-reported questionnaire targeting community-dwelling older adults aged ≥65 years. We used the incidence of depressive symptoms during follow-up as the outcome, the number of remaining teeth (≥20/≤19) as the exposure, and the deterioration of oral function and orofacial appearance (speaking, smiling, and eating) as mediators. We fitted the logistic regression model including confounders and calculated the natural indirect effect (NIE), natural direct effect, and the proportion mediated (PM) by the deterioration of oral functions by applying the causal mediation analysis framework. RESULTS: The mean age of the 8,875 participants was 72.7 years (1SD=5.5) and 48.0% were male. Depressive symptoms developed during follow-up in 11.5% of the participants (n=1,024). The incidence for those with ≤19 and ≥20 remaining teeth was 13.1% and 9.2%, respectively. Total effect of fewer remaining teeth (≤19) on depressive symptoms was OR=1.30 (95%CI=1.12-1.51). Difficulty in speaking (NIE OR=1.03, 95%CI=1.00-1.06, PM=12.4%), problems in smiling (NIE OR=1.04, 95%CI=1.01-1.07, PM=16.9%), and difficulty in chewing (NIE OR=1.05, 95%CI=1.02-1.09, PM=21.9%) significantly mediated the relationship. LIMITATION: Selection bias due to dropout. CONCLUSION: Deterioration of oral function and orofacial appearance were mediating factors of the mechanism for the relationship between tooth loss and the incidence of depressive symptoms.