BACKGROUND: Evidence on the association of the Mediterranean-Dietary Approaches to Stop Hypertension Intervention for Neurodegenerative Delay (MIND) diet with cognitive aging is limited and inconsistent. OBJECTIVES: We examined how the MIND diet is related to cognitive function and its decline among middle-aged and older adults. METHODS: We included 4066 participants with baseline dietary assessment and ≥1 cognitive test from the China Health and Nutrition Survey in 1997, 2000, 2004, and 2006, with a median follow-up of 3 y. The modified MIND diet score (range: 0-12) was calculated based on 9 healthy and 3 unhealthy food groups. Linear mixed-effect models were used to examine the association of adherence to the MIND diet with z-scores of cognitive function and cognitive decline. We also conducted a meta-analysis including our findings and 7 other cohort studies. RESULTS: At baseline, the median MIND diet scores across increasing tertile were 3.0, 4.0, and 5.5, respectively. Participants with higher MIND diet scores had better global cognitive function. The adjusted difference in global cognitive function z-score for every 3-point increment of MIND diet scores was 0.110 [95% confidence interval (CI), 0.060, 0.159, P-trend < 0.001], which was approximately equivalent to being 1 y younger in age. Consumption of nuts, fish, red meats, and tea showed independent positive associations with cognitive function, while fried food consumption exhibited inverse associations. In the meta-analysis of 26,103 participants, one standardized deviation increment of the MIND score was associated with 0.042 (95% CI: 0.020, 0.065) units higher in global cognitive function z-score and 0.010 (95% CI: -0.001, 0.021) units slower in annual cognitive decline. CONCLUSIONS: Our findings suggest that higher adherence to the MIND diet was associated with better cognitive function and potentially slower cognitive decline in later life. Further large-scale observational and interventional studies are warranted to elucidate the cognitive effects of the MIND diet. This meta-analysis was registered at PROSPERO as CRD42022330417.