Purpose: We aimed to identify clinically meaningful Post-Concussion Symptom Scale (PCSS) scores in pediatric concussion patients relative to patients with referral orders and prolonged recovery. Prolonged recovery is defined as greater than 30 days to return-to-learn (RTL) and return-to-play (RTP). Methods: Patients aged 8-17 years reporting to a tertiary care concussion clinic were invited to participate from August 2014 through February 2018 in a prospective cohort study. Participant data collected through electronic medical records included PCSS at initial visit, ordered referrals (ophthalmology, physical therapy, occupational therapy, neuropsychology, child psychology, neurology, or other service), date of injury, and date of RTL/RTP. Dates for RTL and RTP were determined as the day when the patient was able to return to school without accommodations and physical activity without restrictions, respectively. Patients were excluded if they were non-English speaking or had structural abnormalities on neuroimaging. Statistical analysis was restricted to patients who presented to clinic ≤30 days post-injury. T-tests were conducted to compare PCSS scores between patients who returned-to-learn/play before and after 30 days, and between patients who did and did not receive a referral order. Results: Of 486 recruited patients (14.1±2.4 years; 56.2% female), 359 (73.9%) presented to clinic within 30 days from injury. Of the 359, 221 (61.6%) and 230 (64.1%) had documented dates for RTL and RTP, respectively. Median number of days from injury to RTL and RTP was 46 and 52, respectively. Of the 221 patients with RTL dates, 144 (65.2%) returned-to-learn >30 days post-injury. Of the 230 patients with RTP dates, 163 (70.9%) returned-to-play >30 days post-injury. PCSS scores were significantly greater for patients who returned-to-learn >30 days (32.5±28.9 vs. 11.6±8.2, p<0.01) and returned-to-play >30 days (31.9±28.4 vs. 12.3±8.9, p<0.01) from injury. Two hundred thirty-six participants (65.7%) received at least one referral order (Table 1). PCSS scores were significantly greater for patients who received referrals compared to patients who did not (35.9±24.5 vs. 20.8±21.2, p<0.01). Conclusion: To our knowledge, there are no studies identifying clinically meaningful thresholds in PCSS scores for pediatric concussion patients. Our rudimentary analysis suggests patients in our cohort with an initial PCSS score greater than 32 were associated with prolonged number of days to RTL and RTP. Similarly, PCSS scores above approximately 36 were associated with receiving referral orders. Intuitively, increased PCSS scores are associated with greater frequency of referrals and number of days to recovery. Our study suggests potential cutoffs for clinicians to reference when managing severe concussion symptoms. Future studies could provide further threshold analysis with considerations for factors such as sex and history of mental health disorders, as previous literature shows increased reports of symptom severity on PCSS by females and those with a history of anxiety and depression.