Objective: Research on the neurocognitive consequences of sports-related concussion has demonstrated that most athletes return to baseline within 10 days of sustaining a mild traumatic brain injury (mTBI). However, for reasons that remain poorly understood, some athletes recover at a slower rate. The present study examines factors associated with protracted mTBI recovery. Participants and Methods: Five-hundred and seventy-four non-injured college athletes were administered the ImPACT as part of a baseline assessment. Ninety athletes subsequently sustained mTBI during their college tenure, and were re-tested approximately 48 hours postinjury. Athletes received follow-up testing as needed, until a final return-to-play decision was made. For the present analyses, athletes were divided into three groups based on recovery course: 1. athletes with one post-injury evaluation (N=61), 2. those with two evaluations (N=22), and 3. those with three or more evaluations (N=7). Results: One-way ANOVA revealed that performance on the ImPACT at baseline, but not post-concussion, distinguished the groups. The effect of recovery group was significant for the Verbal Memory Composite (F=5.20, p<.01), Visual Memory Composite (F=3.90, p<.05), and Total Symptom Score (F=4.54, p<.05). Athletes who returned to play after their initial post-concussion evaluation exhibited better verbal and visual memory, and reported fewer symptoms, at baseline than those who were re-referred for further post-concussion testing on a second and third occasion. Conclusions: Results suggest that pre-injury factors—symptom reporting and poorer performance on memory tasks—predict a longer recovery time after mTBI. This finding is consistent with theories of cognitive reserve. Theoretical and clinical implications will be discussed.