This study has been done for the purpose of determining whether the positive association between social support and depression is attributable to an overall beneficial effect of support(direct effect) or to a process of support protecting persons from adverse effects of stressors such as life events, pain or physical disability (buffering effect). The sample consisted of 214 patients who were identified as the rheumatoid arthritis. The instruments used in this study were Depression Scale(CES-D), Perceived Social Support Scale, Life Events Questionnaire, AIMS Pain Scale, and Physical Disability Scale. The data were analysed by the use of t-test, ANOVA, Pearson Correlation Coefficient and Stepwise Multiple Regression. The results of this study are summerized as follows : 1. The 1st hypothesis, "The higher the life events degree, the higher the depression degree" was supported(r=.49, P=.0001). 2. The 2nd hypothesis, "The higher the pain degree, the higher the depression degree" was supported(r=.44, P=.0001). 3. The 3rd hypothesis, "The higher the physical disability degree, the higher the depression degree" was supported (r=.46, P=.0001). 4. The 4th hypothesis, "The higher the social support degree, the lower the depression degree" was supported(F=84.52, P=.0001). 5. The 5th hypothesis, "There will be different in the relationship between the degrees of life events and depression according to social support degree" was rejected (F=.29, P=.5928). 6. The 6th hypothesis, "There will be different in the relationship between the degrees of pain and depression according to social support degree" was supported (F=3.19, P=.0755). 7. The 7th hypothesis, "There will be different in the relationship between the degrees of physical disability and depression according to social support degree" was supported(F=5.69, P=.018). 8. The predictive variables for depression were the degrees for social support, life events, pain, and physical disability. 9. The depression degree showed a inverse correlation with social support degree (r=-.56, p=.0001). The social support degree showed a inverse correlation with the degrees of life events(r=-.22, p=.0007), pain(r=-.18, p=.0069) and physical disability(r=-.15, p=.0293). 10. The depression degree showed significant differences in the variables of sex (t=2.26, p=.025), educational level(r=.189, P=.006) and the number of treatment method (r=.201, P=.003). In conclusion, it was found that social support had the direct effect on depression and the buffering effect in each relationship between degrees of pain or physical disability and depression in patients with rheumatoid arthritis. So the researcher thinks that it is effective that nurses should provide these patients with social support to reduce depression in cases of having severe pain or physical disability.