Superficial digital flexor tendon (SDFT) tendinopathy is an important musculoskeletal problem in horses. The study objective was to validate an ultrasonographic scoring system for SDFT injuries. Ultrasonographic images from 14 Thoroughbred racehorses with SDFT lesions (seven core; seven diffuse) and two controls were blindly assessed by five clinicians on two occasions. Ultrasonographic parameters evaluated were: type and extent of the injury, location, echogenicity, cross-sectional area and longitudinal fibre pattern of the maximal injury zone (MIZ). Inter-rater variability and intra-rater reliability were assessed using Kendall’s coefficient of concordance (KC) and Lin’s concordance correlation coefficient (LC), respectively. Type of injury (core vs. diffuse) had perfect inter/intra-rater agreement. Cases with core lesions had very strong inter-rater agreement (KC ≥0.74, P<0.001) and intra-rater reliability (LC ≥0.73) for all parameters apart from echogenicity. Cases with diffuse lesions had strong inter-rater agreement (KC ≥0.62) for all parameters, but weak agreement for echogenicity (KC=0.22); intra-rater reliability was excellent for MIZ location and fibre pattern (LC ≥0.82), and moderate (LC ≥0.58) for cross-sectional area and number of zones affected. This scoring system was reliable and repeatable for all parameters, except for echogenicity. A validated scoring system will facilitate reliable recording of SDFT injuries and inter-study meta-analyses.