Objective: The primary objectives were to establish Cantonese versions of several measures of executive functions and to determine which measures differentiated clinical versus nonclinical participants in a Chinese society. Method: Participants included nonclinical community residents of Macao, China (n = 54; M age = 37) recruited through a neighborhood community center as well as residents with an identified (via medical chart review) psychiatric disorder (primarily Schizophrenia, mood disorder, or substance abuse/dependence) but no language disorder (n = 49; M age 36) recruited through a community mental health center and a residential substance abuse treatment facility. All were native Cantonese speakers. Participants received a small honorarium (the equivalent of approximately $10USD) and were administered a group of standardized tests that included measures of executive functions (e.g. a Semantic Fluency test and Pair Cancellation from the Woodcock Johnson III) and measures of processing speed (Visual Matching 2 and Decision Speed from the Woodcock Johnson III). Results: After controlling for age and education, groups did not differ significantly in processing speed. The clinical group generated significantly fewer words on the Semantic Fluency test (F(1,98) = 10.23, p = .002, partial ƞ2 = .10) and worse performance on Pair Cancellation (F(1,100) = 19.97, p < .001, partial ƞ2 = .17), indicating effect sizes in the medium range. Conclusion: This study contributes to the limited research on neuropsychological test performance in Cantonese-speaking Chinese and suggests Semantic Fluency (“fruits” and “animals” categories) and Pair Cancellation could be useful in identifying executive function deficits in psychiatric patients in (or from) Cantonese-speaking areas.