Abstract Purpose Removal from play after a concussion is a standard requirement across sports in order to reduce risk of re-injury. This study examines the effect of ongoing play on early symptom status and overall length of recovery. Methods 191 adolescents aged 11-19 (mean age=15.3± 1.6, 57% male) presenting within 30 days of a concussion completed the Post-Concussion Symptom Inventory-2 (PCSI-2). Participants were recovered when their PCSI-2 Total Retrospective Adjusted Post-Injury Difference (RAPID) scores fell below the psychometrically established cut-off for minimal symptoms. ANOVAs compared initial symptom status and length of recovery (LOR) between those who continued play (CTP) and who were removed from play (RFP) after injury, by sex. Results No difference in initial symptom level was found between individuals who CTP versus those RFP (PCSI-2 RAPID mean 24.7 vs. 22.1, p=.41). A marginal difference was found in LOR between CTP versus RFP (31.7 vs. 27.2 days, p=.088). Marginally higher symptom levels were found for females vs. males (mean=26.4, 20.4, respectively, p=.055). Females took longer to recover than males (33.7 vs. 25.2 days, p=.001, η2=.05). There was no interaction between CTP by sex on symptom levels (p=.92) or on LOR (p=.10), after accounting for CTP and sex alone. However, females who CTP had longer LOR compared with females who were RFP (38.0 vs 29.3 days, p=.029). Conclusions Findings show marginal overall difference in symptom status but not LOR based on immediate removal from play. Females who CTP had longer recovery suggesting a possible sex-specific interaction effect of ongoing play after concussion.