OBJECTIVE: To cross-culturally adapt and test the FRAIL scale in Chinese community-dwelling older adults. DESIGN: Cross-sectional study. METHODS: The Chinese FRAIL scale was generated by translation and back-translation. An urban sample of 1235 Chinese community-dwelling older adults was enrolled to test its psychometric properties, including convergent validity, criterion validity, known-group divergent validity, internal consistency and test-retest reliability. RESULTS: The Chinese FRAIL scale achieved semantic, idiomatic, and experiential equivalence. The convergent validity was confirmed by statistically significant kappa coefficients (0.209-0.401, P < .001) of each item with its corresponding alternative measurement, including the 7th item of the Center for Epidemiologic Studies-Depression Scale, the Timed Up and Go test, 4-m walking speed, polypharmacy, and the Short-Form Mini Nutritional Assessment. Using the Fried frailty phenotype as an external criterion, the Chinese FRAIL scale showed satisfactory diagnostic accuracy for frailty (area under the curve = 0.91). The optimal cut-point for frailty was 2 (sensitivity: 86.96%, specificity: 85.64%). The Chinese FRAIL scale had fair agreement with the Fried frailty phenotype (kappa = 0.274, P < .001), and classified more participants into frailty (17.2%) than the Fried frailty phenotype (3.9%). More frail individuals were recognized by the Chinese FRAIL scale among older and female participants than their counterparts (P < .001), respectively. It had low internal consistency (Kuder-Richardson formula 20 = 0.485) and good test-retest reliability within a 7- to 15-day interval (intraclass correlation coefficient = 0.708). CONCLUSIONS: The Chinese FRAIL scale presents acceptable validity and reliability and can apply to Chinese community-dwelling older adults.