Background: Chinese nationwide norms of the Mini-Mental State Examination (MMSE) have not been established despite its wide use. Objective: To obtain norms for the MMSE based on age, gender, education, and rural or urban residences and to determine the optimal cut-off points of the MMSE in elderly Chinese. Methods: A cross-sectional study was conducted in Chinese community residents aged 65 years or over selected by cluster random sampling. The MMSE was administered to 9,629 subjects (7,110 cognitively normal, 2,024 with mild cognitive impairment, and 495 with dementia). The demographic influences on MMSE scores were investigated and the norms were established considering those factors. Receiver operating characteristic (ROC) analysis was used to determine the optimal cut-off points. Results: Years of education (standardized beta = 0.399), rural residence (standardized beta = -0.261), age (standardized beta = -0.198), and being female (standardized beta = -0.101) had significant effects on MMSE scores (p < 0.001). Accordingly, we presented the demographic-stratified normative data for the MMSE. The optimal cut-off points for dementia screening were 16/ 17 for illiterate (sensitivity 87.6% and specificity 80.8%), 19/ 20 for individuals with 1-6 years of education (sensitivity 93.6% and specificity 92.7%), and 23/ 24 for individuals with 7 or more years of education (sensitivity 94.3% and specificity 94.3%). Conclusion: We provide the age-, gender-, education-, and residence-specific reference norms for the MMSE derived from an investigation of a large-scale, multicenter, nationwide representative Chinese elderly population. It could be of great improvement for the use of the MMSE in dementia screening in Chinese elderly population.