The current study aimed to evaluate varying data integration procedures and their effects on the classification accuracy of Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT). Data were collected from an independent secondary school that included students in grades 9 to 12. The study examined 300 consecutive ImPACT score reports generated by secondary school students between 2010 and 2015. To appraise ImPACT's utility as a serialized measure, standardized regression-based equations were formulated to compute reliable change index scores. Discriminant function analyses (DFAs) consisting of varying combinations of ImPACT composite scores were conducted and their accuracy was compared to that produced by the standard interpretive procedure. Varying combinations of scores produced on Verbal Memory, Visual Memory, Visual Motor Speed, Reaction Speed, and the Postconcussion Symptom Scale were included in analyses. DFAs yielded sensitivities ranging from 31% to 49%, specificities from 88% to 95%, positive predictive values (PPVs) from 61% to 83%, and negative predictive values (NPVs) from 67% to 75%. Conversely, the standard interpretive procedure yielded a sensitivity of 73%, specificity of 43%, PPV of 45%, and NPV of 72%. The standard interpretive procedure produced a higher sensitivity than the DFAs; however, its PPV did not exceed chance levels. Conversely, DFA equations produced superior PPVs; however, their sensitivity hovered around 50%, leaving a substantial proportion of individuals with concussion undetected. Cognitive composite scores did not appear to offer significant incremental utility in relation to symptom self-report. Base rate conditions and psychometric factors appeared to contribute to ImPACT's limited classification accuracy.