BACKGROUND: There is increasing evidence of causal links between poor mastication and cognitive impairment, but possible effects of dietary hardness, which clearly affects mastication, on cognitive function are unknown. OBJECTIVE: We investigated the hypothesis that hardness of the habitual diet would be associated with cognitive function among older Japanese adults. METHODS: The subjects of this cross-sectional study were 635 Japanese community-dwelling people aged 69-71 years. The masticatory muscle activity required for the habitual diet was used to determine dietary hardness. Consumption of 38 foods was assessed by a validated, brief-type, self-administered diet history questionnaire. A published database was then used to estimate the masticatory muscle activity involved in the ingestion of these foods. The Japanese version of the Montreal Cognitive Assessment (MoCA-J) was used for the measurement of cognitive function. RESULTS: The principal contributors to dietary hardness were cooked rice (28.0%), green leafy vegetables (5.1%), dried fish (4.9%), and pork and beef (4.6%). There was a positive association between dietary hardness and MoCA-J score that was robust to adjustment for potential confounders (MoCA-J score per 100-unit increase in dietary hardness: β = 0.83 [95% CI: 0.08, 1.59], P = 0.03). These results did not change materially even after exclusion of subjects who reported substantial changes in their diet for any reason (β = 0.94 [95% CI: 0.02, 1.86], P = 0.04). CONCLUSION: This preliminary cross-sectional study suggests that dietary hardness might have a beneficial effect on cognitive function in older Japanese people. Further prospective studies with more accurate measurements are needed to confirm this finding.