Based on Messick's (1995) framework of construct validity and responses from 521 teachers in the United States and 580 teachers in China, this study provided evidence of the generalizability aspect of construct validity for the Classroom Appraisal of Resources and Demands (CARD) questionnaire originally developed in the United States. Confirmatory factor analysis, factorial invariance testing, and differential item functioning were employed to examine the systematic variation of interpretation by the Chinese and US elementary school teachers with respect to the 65 items in the CARD. Results suggested that responses to CARD follow the same structure from both Chinese and US teachers, and that 83% of items showed no difference. Differences in Chinese and US teachers' interpretations of certain items were discussed with implications to school psychology, school counseling, teacher education, educational administration, and policy research.