ObjectiveTo develop a whole joint scoring system, the Anterior Cruciate Ligament OsteoArthritis Score (ACLOAS), for magnetic resonance imaging (MRI)-based assessment of acute anterior cruciate ligament (ACL) injury and follow-up of structural sequelae, and to assess its reliability. DesignBaseline and follow-up 1.5 T MRI examinations from 20 patients of the KANON study, a randomized controlled study comparing a surgical and non-surgical treatment strategy, were assessed for up to six longitudinal visits using a novel MRI scoring system incorporating acute structural tissue damage and longitudinal changes including osteoarthritis (OA) features. Joint features assessed were acute osteochondral injury, traumatic and degenerative bone marrow lesions (BMLs), meniscus morphology and extrusion, osteophytes, collateral and cruciate ligaments including ACL graft, Hoffa-synovitis and effusion-synovitis. Cross-sectional (baseline) and longitudinal (all time points and change) intra- and inter-observer reliability was calculated using weighted (w) kappa statistics and overall percent agreement on a compartmental basis (medial tibio-femoral, lateral tibio-femoral, patello-femoral). ResultsAltogether 87 time points were evaluated. Intra-observer reliability ranged between 0.52 (baseline, Hoffa-synovitis) and 1.00 (several features), percent agreement between 52% (all time points, Hoffa-synovitis) and 100% (several features). Inter-observer reliability ranged between 0.00 and 1.00, which is explained by low frequency of some of the features. Altogether, 73% of all assessed 142 parameters showed w-kappa values between 0.80 and 1.00 and 92% showed agreement above 80%. ConclusionsACLOAS allows reliable scoring of acute ACL injury and longitudinal changes. This novel scoring system incorporates features that may be relevant for structural outcome not covered by established OA scoring instruments.