Numerous research reports have substantiated the role of stressful life events in relation to the onset of health changes. The relationship tends to hold across different age groups. Theoretically, adolescence has been considered a developmental crisis period of great stress, impoverished coping skills and high vulnerability to biological, social and psychological demands. The research problem addressed by this study was to examine the relationships between stressful life events and health symptom patterns, and the effect of two variables, coping and social, support, theoretically considered to mediate the relationship between stress and health symptoms in adolescents. The following five hypotheses were tested in this research : 1. Health symptoms are positively related to stressful life events in adolescents, 2. Health symptoms are negatively related to coping in adolescents, 3. Health symptoms are negatively related to social support in adolescents, 4. When coping is controlled, the relationship between health symptoms and stressful life events will decrease, and 5. When social support is controlled, the relationship between health symptoms and stressful life events will increase. The study subjects consisted of 1090 high school students of the metropolitan city of Seoul. The following sampling procedure was used : 1. Of the 169 high schools in nine school administrative districts in the city, a proportional sample of ten schools was selected. 2. One class from each of the freshman and sophomore was randomly selected and all the students who were in the sampled class were used as the study sample. The study was limited to freshman and sophomore adolescents, aged 15 to 18(mean＝16.6). Of the 1090 subjects 688(63%) were boys and 402(37%) were girls. An Adolescent Inventory of Stressful Life Events, a Health Symptom Questionnaire and an Adolescent Coping Inventory were adapted for this study. The Norbeck Social Support questionnaire was utilized to collect the data on perceived social support. Five high school teachers in the areas of school health and counselling reviewed the items of each questionnaire for content validity. A pilot study was undertaken to ascertain reliability. Fifty three high school students responded to the questionnaires and gave their opinions on the items. For stressful life events, health symptoms, coping, and social support, the Cronbach's alpha's on the study were .70, .94, .77, and .76, respectively. Research assistants attended all the sampled classes with the school proctor to explain the purpose and procedures of the study to the students. The questionnaires along with a ballpoint pen were distributed to the students who were asked to complete each item. The research assistants left the ballpoint pen with the students as a gift for their cooperation. An average of 50 minutes was required to complete the questionnaires. Using an SPSS, the first, three hypotheses were tested using Gamma, a measure of association for ordinal variables. Partial gamma was used to test the fourth and fifth hypotheses. Patterns of elaboration described by Babbie were selected to interpret the relationship of the three variable analyses. The significance of gamma was determined by Chisquare at a .05 level of significance. There was a positive relationship between health symptoms and stressful life events(Gamma＝.35, p＝.000). Thus the first hypothesis was supported. Unexpectedly, coping was positively related with health symptoms(Gamma＝.13, p＝.000). That is, the higher the coping levels, the greater number of health problems. The third hypothesis, the higher the level of social support, the fewer the health symptoms, was not accepted in this adolescent study group. When coping was controlled, under the condition of low coping the association between health symptoms and stressful life events increased significantly to a partial gamma of .39, and under the condition of high coping it was .30. According to the elaboration model, when one partial relationship is the same or greater than the original and the other is smaller, the control variable should be considered to be specifying the conditions. When social support was controlled the relationship between stressful life events and health symptoms increased under the condition of low social support, but with high social support, the relationship decreased. Both partial gamma were statistically significant at .05 level(.43 and .26 relatively). It can be interpreted that stressful life events are strongly and positively related to health symptoms under the condition of low social support, however this relationship can not be expected with high social support. Thus, the last two hypotheses were conditionally sustained. In this study, the relationships between stressful life events and health symptoms, and the specified me diating roles of coping and social support were found to have statistical interaction. This finding supports the theoretical position of this study. It suggests that stressful life events would create high susceptability to biological social and psychological health symptoms and coping and social support buffering the relationship between stressful life events and health symptom. The findings of this study have implications for nursing practice. When adolescents are confronted with non-developmental life events that are perceived as stressful, nurses should recognize the evidence of the stress-buffering effect of coping and social support on health symptoms and utilize the diverse sources of social support that are readily available to adolescents.