Objective: Predicting recovery duration from a sports-related-concussion within initial days post-trauma using cognitive testing has proven difficult. In this study, we determined which test scores from initial post-trauma evaluation best predicted a typical (7–12 days) versus prolonged recovery and return to play. Method: The Standardized Concussion Assessment Tool 3 (SCAT-3) was administered to 34 (62% male) NCAA Division II student athletes, aged 18 to 27 (M = 20.20, SD = 1.59), as a pre-competition baseline and 24–72 hours post-trauma. Presence and recovery from concussions were determined by review of symptom presentation and assessment data by a licensed neuropsychologist, team physician, and athletic trainer. Results: Average recovery time was 50 days (SD = 68.03, Median = 16.5, Range = 7–226). The average change score on the SAC portion of the SCAT-3 was 2.26 points lower on post-trauma than baseline. Stepwise regression analysis found SAC change score was the only significant predictor of recovery time (R = −0.403, p = .018) when the difference was ≥2. Finally, a t-test found SAC change scores differed significantly between those with typical versus prolonged recovery time (p = .022). Changes in total symptoms, symptom severity, and balance were not predictive of prolonged recovery. Conclusion(s): Decrease in SAC score from baseline was predictive of a prolonged recovery time, even when the SAC was administered up to 72 hours post- trauma. Such changes can alert the sports medicine team that an athlete's risk for delayed recovery is significant so prospective supports may be implemented early in the recovery process.