Objective We aim to study the effect of persistent post-concussive symptoms on cognitive performance of elite athletes, based on the hypothesis that persistent symptoms could impair athletes’ cognitive performances and functional outcomes.Design Prospective observational and longitudinal case control study.Setting University research centre, single centre.Participants 7 athletes with persistent symptoms (>2 weeks, assessed at median 31 days post-concussion), 10 athletes with brief symptoms (<3 days, assessed within a week, median 3.5 days), and 9 age-matched healthy volunteers. Inclusion criteria: male/female athletes (aged 18–40) in contact sports who sustained concussion(s) recently, being symptomatic with normal neuro-radiological findings at enrolment.Intervention Independent variables: post-concussive symptom scores and duration of symptoms.Outcome measures Multi-dimensional scores from Sport Concussion Assessment Tool III, Immediate Post-concussion Assessment and Cognitive Testing, WAIS symbol search and digit span backward tests.Main results Patients with persistent symptoms scored less in the symbol search test than those with brief symptoms and the healthy volunteers (grouping effect p=0.006; post hoc tests p=0.039 and 0.004, respectively). In all 17 patients, symptom scores correlated with reaction time (ρ=−0.61, p=0.009), and visual motor speed correlated with symbol search scores (ρ=0.59, p=0.013). 6 patients who completed follow-up assessments when symptom-free showed improvement in their symbol search scores (p=0.042) and visual memory (p=0.028).Conclusions Concussed athletes with persistent symptoms performed poorly in visual perception/analysis and information processing. Persistent post-concussive symptoms could affect athletes’ visual memory, perception/analysis, reaction time and information processing speed, thus precipitating them to further injuries should they return to play prematurely.