OBJECTIVES: Study 1: To determine the interrater agreement on the Multiattribute Health Status Classification (MAHSC) for brain-injured children. Study 2: To determine the outcome of severe childhood traumatic brain injury (TBI) by comparing three measures: MAHSC, Functional Independence Measures (FIM/WeeFIM), and the Glasgow Outcome Scale. Designs: Study 1: Clinic recruitment of parents of patients. Study 2: Surveillance follow-up of an inception cohort. Settings: Study 1: The Brain Injury Clinic, Glenrose Rehabilitation Hospital, Edmonton, Canada. Study 2: Pediatric Intensive Care Unit, University of Alberta Hospital. PATIENTS: Study 1: Two physiatrists and parents of 50 children (5-18 yrs, 54% boys) independently completed the survey. Study 2: From a cohort of 51 patients (3-17 yrs, 69% boys, 6 deaths) consecutively admitted to the pediatric intensive care unit in 1995 and 1996 with severe TBI (Glasgow Coma Score </= 8 within the first 24 hrs postinjury), parents of all survivors (71% boys) completed outcome measures at 6-12 months postinjury. MEASUREMENTS AND MAIN RESULTS: Study 1: The interrater agreement exceeded 70% for attributes of sensation, mobility, cognition, self-care, and general health. Study 2: Of 45 survivors, 34 (76%) had a "good recovery" on the Glasgow Outcome Scale, 16 (36%) had normal scores on the FIM/WeeFIM, and only 8 (18%) had normal attributes on the MAHSC. Correlations of measures were Glasgow Outcome Scale and MAHSC, -.73; Glasgow Outcome Scale and FIM/WeeFIM,.64; and MAHSC and FIM/WeeFIM, -.63. Sensitivity and specificity from acute injury predictors for the Glasgow Outcome Scale were 88% and 91%, respectively; for MAHSC 75% and 70%; and for FIM/WeeFIM 63% and 75%. CONCLUSIONS: The MAHSC has a high interrater reliability after brain injury and is a useful parent-report surveillance tool to audit outcome after severe TBI. It identified problems not addressed by the Glasgow Outcome Scale or FIM/WeeFIM. Most children with severe TBI have adverse outcomes.