Objective: Assessment of the impact of concussion in children and adolescents includes both symptom survey and neurocognitive assessment. Computerized measures of neuropsychological functioning can be useful in detecting subtle improvements in neurocognitive functioning over time. The objective of this paper is to examine the utility of a battery of computerized measures of memory, reaction time, and processing speed in capturing the nature and trajectory of the recovery process of children and adolescents who have sustained a concussion. Participants and Methods: Participants included 122 children and adolescents ages 11 to 19 referred to a hospital-based concussion clinic over a 4 year period. Most participants were seen for 2-3 visits, with the first visit typically within a month of the injury. ImPACT was administered at each visit (using alternate forms). Parent and patient symptom inventories were also included at each visit. Results: After accounting for variance in outcome due to the effects of time (linear and quadratic change), we developed growth curve models accounting for significant proportions of remaining variance in ImPACT composite scores (verbal and nonverbal memory, reaction time, and visual motor speed). Improvements over time were fairly linear. Gross categories of cause of injury accounted for far more variance in outcome over time than such traditional factors as the presence of LOC, anterograde amnesia or retrograde amnesia. Baseline symptom reports explained only minimal amount of variance in outcome scores. Conclusions: A multi-dimensional assessment model that includes both symptom report and computerized neuropsychological evaluation provides a comprehensive analysis of recovery trajectories after child and adolescent concussion.