Objective  To revise the Connor-Davidson resilience scale (CD-RISC) and study its reliability and validity in Chinese soldiers. Methods  Reliability and validity were detected in 1 800 Chinese soldiers with Chinese version CD-RISC

Eysenck personality questionnaires (EPQ)

self-report symptom checklist 90 (SCL-90)

simplified coping style questionnaires and post-traumatic stress disorder checklist-civilian v

ersion (PCL-C)

respectively. Results  The exploratory factor analysis showed that CD-RISC was composed of 15 items

including the tenacity and adaptability of soldiers. The accumulated variance  explanation rate was 43.148%. The CD-RISC factors and total score were positively correlated with positive coping style and resilience in simplified coping style questionnaires and EPQ

but negatively correlated with anxiety

depression and somatolization in SCL-90

and re-experience

avoidance and high waking in PCL-C

and neural dimention in EPQ (

P

<

0.01）. The anxiety

depression and somatolization in SCL-90

and re-experience

avoidance and high waking in PCL-C were significantly higher in soldiers with a low CD-RISC score than in those with a high CD-RISC score（

P

<

0.01）. The consistency coefficient of total resilience scale and Cronbach’s α of each factor was 0.767 to 0.816. The correlation coefficient between factors and the sum of resilience scale was 0.640 to 0.913（

P

<

0.01）. Conclusion  The reliability and validity of Chinese version CD-RISC comprised of 15 items are rather good in Chinese soldiers. The resilience level can forecast individual’s mental health state after stress. Positive coping style and personality characteristics of extraversion improve individual resilience.