Objectives:The primary purpose of this study is to understand the psychopathology of the victims of school violence in terms of early psychosis. By doing this, the early detection of psychosis among the victims is possible, and early detection may lead to early intervention. Methods:Two-thousand and nine-hundred seventy two students from 16 middle schools in Seoul were asked to fill out questionnaire comprised of popularity and intellectual and school status of Piers-Harris Children's Self Concept Inventory, Symptom Check List-90-Revised, and Ostracism Scale. The subjects whose scores upon Ostracism Scale were higher than average by two standard deviation were labeled as ‘Repelled and Isolated group', and subjects whose scores on popularity were significantly lower than average and whose scores on psychoticism of SCL-90-R were higher than average were defined as 'tentative early psychosis group'. Odds ratios were calculated from the numbers of subjects with and without high psychoticism scores and high ostracism scores. On the subjects of 'tentative early psychotic group', we examined every clinical characteristic and conducted correlation analysis and regression analysis in order to find out the risk factors and to construct theoretical model that explains the psychoticism scores. Results:The results were as follows:1) Total 157(5.3%) subjects were rated significantly higher on ostracism scale, and among them, 47 subjects(29.9%) were rated significantly higher than average on psychoticism scale, while only 50 subjects among 2,135 students who were rated within normal range showed significantly higher score on psychoticism scale. Odds ratio for psychotic group of isolated group were 17.82 and it was statistically significant. 2) Forty-seven subjects(31 boys, 16 girls) who were rated as they were unpopular and rejected from peers had significantly higher psychoticism scores. They were not significantly different from simply high psychoticism subjects in anxiety, social anhedonia scale, magical thinking, obsessivecompulsive symptoms, phobic anxiety, psychoticism, somatization, but showed higher ostracism scores and paranoid tendencies. Among school violence victims, who rated themselves unpopular and showed higher psychoticism scores, the psychoticism scores were mainly explained by anxiety, depression, hostility, interpersonal sensitivity, obsessive-compulsive symptoms, paranoid tendency, somatization scales($r^2=0.93$). Conclusion:Thus, it can be concluded that the subjects with higher ostracism score have the substantially high risk for psychosis development. By these results, we propose that school violence victims with anxiety, depression, hostility, interpersonal sensitivity, obsessive-compulsive symptoms, paranoid tendency, somatization should be tested individually considering school adjustment, attentional deficit, concept formation problems.