Background Using functional near-infrared spectroscopy (fNIRS) as an objective diagnostic tool, we aimed to (1) compare fNIRS measurements in adult and youth male rugby players against controls over a playing season, and 2) document the fNIRS changes that occur in concussed rugby players during the England Rugby Football Union Graduated Return-To-Play programme (GRTP). Sixty-seven participants (rugby = 41 (26 adults: 27.5 ± 4.4 years; 15 youth: 16.5 ± 0.6 years; control = 26 (11 adult: 30.5 ± 5.2 years; 15 youth: 16.9 ± 0.4 years) completed fNIRS assessments at pre, mid and end-season. Eight players (five youth, three adult) sustained concussions, and completed fNIRS and the Graded Symptom Checklist from the Sport Concussion Assessment Tool version 5 (SCAT5) assessment throughout the GRTP period. Mixed linear models were utilised to assess the effect of group and time on fNIRS measures of oxyhaemoglobin (∆O2Hb) and deoxyhaemoglobin (∆HHb) during performance tasks. Typical Error (TE) i.e., normal biological fluctuation and measurement error, was calculated to identify ‘cut-off’ thresholds for identifying effects of concussion. Results There were significant differences in fNIRS indices over time in adult and youth groups (p textless 0.05) but no significant differences between rugby and control groups (p textgreater 0.05). Seven out of eight (87.5%) concussed players showed changes greater than TE during the GRTP period for both ∆O2Hb and ∆HHb during performance tasks and these players’ ∆O2Hb profiles had not returned to within ‘normal’ levels within the GRTP period. All players’ symptom severity and number returned to normal within the GRTP period. Conclusion Current GRTP protocols alone are problematic and there is a need for a more individualised approach to concussion management, utilising objective biomarker tools such as fNIRS.