Purpose: Rapid and accurate assessment of impaired consciousness is very important, especially for critically ill patients. Therefore, the Glasgow coma scale (GCS) has been widely adopted in the assessment of adult and pediatric comas. But the GCS should not be used for small children who show various normal responses according to their developmental stage. Therefore, the objectivity of pediatric coma measurements was verified by applying the modified GCS to patients in a pediatric intensive care unit (PICU) at a university hospital in Korea. Methods: The level of consciousness was evaluated for 200 cases who were admitted to a PICU from July 4, 2009 to September 18, 2009 and could show a verbal response, with our modified GCS for children. In addition, and we confirmed the frequency analysis and Kappa statistics with SPSS/WIN 17.0. Results: Kappa statistics which show inter-observer reliability were very good for all components (eye opening, verbal, and motor score) and was good for total GCS scores. Conclusion: It is concluded that our modified GCS is reliable. Therefore, reliable assessment for the level of consciousness is clinically practicable if enough training is supported.