Background: Given the high rates of suicide in the adolescent population and the reluctance of this population to seek help, developing proactive and effective strategies to timely detect individuals at high risk for suicide in non-clinical contexts is a worldwide recognized need. A series of brief self-report questionnaires have been developed for this purpose, however there are few studies providing evidence on their capability to accurately classify suicidal risk levels in specific populations. One of the instruments frequently used to evaluate suicide risk is the Columbia-Suicide Severity Rating Scales (C-SSRS). The goal of this study is to provide psychometric evidence about the accuracy of the Suicidal Ideation subscale (SI) of the C-SSRS to classify suicidal risk levels in a sample of Chilean adolescents using Item Response Theory (IRT). Methods and materials: Through the two parameter logistic model (2-PLM), we analyzed the capability of a self-report questionnaire addressing suicidal ideation (SI) to differentiate and classify participants according to their SI severity levels. We tested two main parameters: difficulty (localization) and discriminating power of 6 items extracted and adapted from the Columbia-Suicide Severity Rating Scales (C-SSRS). We administered this questionnaire to a general sample of 1645 adolescents aged 13 to 18. Results: Our results show that the items differentiate symptoms addressing suicidal thoughts according to their severity, providing an accurate classification of the SI risk level. Conclusions: These findings support the usage of the C-SSRS in Chilean adolescents. Further research is needed to test its predictive value in different populations.