OBJECTIVE: The objective of this study was to determine if computerized neurocognitive testing (Immediate Post-Concussion Assessment and Cognitive Testing [ImPACT]) in the emergency department (ED) can be used as a prognostic tool to detect young athletes at risk of having protracted concussive symptoms. METHODS: This was a prospective cohort study of athletes aged 11 to 18 years who presented to an ED less than 24 hours after sustaining a sports-related concussion. ImPACT was administered in the ED, and performance was categorized as "poor" if the athlete had 3 (of 4) or greater low domain scores. Participants completed the Post-Concussion Symptom Scale (PCSS) in the ED and by phone at 1 and 2 weeks after injury. Athletes were symptomatic if their PCSS score was more than 6 in males and more than 8 in females. RESULTS: One hundred nine patients were enrolled; 60% and 36% remained symptomatic at 1 and 2 weeks after injury, respectively. "Poor" ImPACT performance was not particularly useful in predicting athletes with protracted symptoms (at 1 week: positive predictive value, 70.8%; negative predictive value, 43.5%; at 2 weeks: positive predictive value, 47.8%; negative predictive value, 68.9%). In bivariate analysis, a higher ED PCSS score was associated with protracted symptoms (at 1 week: odds ratio, 1.1 [confidence interval, 1.0-1.1]; at 2 weeks: odds ratio, 1.0 [confidence interval, 1.0-1.1]). CONCLUSIONS: Computerized neurocognitive testing in the ED has limited usefulness in predicting protracted symptoms. Total acute symptom burden may be a useful prognostic tool in the ED evaluation of concussed young athletes, yet further research is necessary.