To determine differences in prevalence and diagnostic accuracy of MRI findings between asymptomatic athletes and athletes with longstanding groin pain. One hundred twenty-three adult male athletes were approached with 85 consecutive athletes recruited. Group 1 (symptomatic, n = 34) athletes referred for longstanding groin pain (insidious onset, > 3 weeks duration). Group 2 (control, n = 51) athletes referred for injuries remote from the pelvis and no groin pain in the last 12 weeks. All referrers completed a clinical examination proforma documenting absence or presence of pelvis and hip abnormality. All patients completed the Copenhagen Hip and Groin Outcome Score (HAGOS) questionnaire and underwent a 3T MRI groin and hip protocol. MRIs were scored independently by two musculoskeletal radiologists blinded to clinical details. Statistical analysis was performed to evaluate associations between MRI findings, inter-reader reliability, clinical examination and HAGOS scores. Pubic body subchondral bone oedema, capsule/aponeurosis junction tear and soft tissue oedema were more prevalent in the symptomatic group (p = 0.0003, 0.0273 and 0.0005, respectively) and in athletes with clinical abnormality at symphysis pubis, adductor insertion, rectus abdominis, psoas and inguinal canal (p = 0.0002, 0.0459 and 0.00002, respectively). Pubic body and subchondral oedema and capsule/aponeurosis tear and oedema significantly correlated with lower (worse) HAGOS scores (p = 0.004, 0.00009, 0.0004 and 0.002, respectively). Inter-reader reliability was excellent, 0.87 (range 0.58-1). Symphyseal bone spurring, disc protrusion and labral tears were highly prevalent in both groups. Clinical assessment and MRI findings of pubic subchondral bone oedema and capsule/aponeurosis abnormality appear to be the strongest correlators with longstanding groin pain.