Objective: To evaluate psychometric properties of a Chinese version of Adult Attachment Questionnaire (AAQ3.1). Methods: Score s on the AAQ3.1 were c ompared among 152 students,114 students' parents and 37 patients. Results: In ternal consistency in terms of item-total correlations was found to be high, but somewhat lower than what was found in the West. High discriminative power was f ound for differentiating between clinical and non-clinical samples, as well as b etween adult and adolescent samples. Significant correlation was found between husbands and wives. Conclusion: AAQ3.1 have acceptable re liability and validity for evaluating attachment of Chinese adults.