Objective:To evaluate the test-retest reliability and validity of Chinese version of World Health Organization Composite International Diagnostic Interview version 3.0(CIDI-3.0)by community-based study.Methods:Among 202 subjects from Dalian city,with the clinician-administered Structured Clinical Interview for DSM-IV(SCID),102 patents were diagnostic as mood disorder,anxiety disorder,schizophrenia or psychotic disorder and so on.All of the patients and the other 100 subjects without mental disorders as the control group were interviewed blindly by CIDI-3.0 to test the validity of CIDI-3.0.Ten patients among them were interviewed twice independently in a 7-day interval to evaluate the reliability of CIDI-3.0.Results:(1)For the screen section,the sensitivity values of different mental disorders ranged from 60.4% to 93.1%,while the specificity values from 33.6% to 92.7%.The positive predictive values were from 60.1% to 95.1%,and the negative predictive values were from 68.1% to 93.7%.(2)For different mental disorders,the specificity values ranged from 97.1% to 98.9%,while the sensitivity values were from 33.3% to 70.3%.Positive predict values were from 66.7% to 95.7%,and negative predictive values were from 87.7% to 95.4%.(3)The consistency was 0.78 in any mental disorder.(4)For test-retest reliability,kappa values ranged from 0.737 to 1.0.Conclusion:By clinical reappraisal,the Chinese version of CIDI-3.0 has satisfied validity and reliability.The screen section has high sensitivity,while the diagnostic sections have high specificities.That indicates that CIDI-3.0 is acceptable as a validated instrument for community survey on mental disorders.