Abstract Background Injury to the supraspinous ligament (SSL) is reported to cause back pain in the horse. The diagnosis is based on clinical examination and confirmed by ultrasonographic examination. The ultrasonographic appearance of the supraspinous ligament has been well described, but there are few studies that correlate ultrasonographic findings with clinical pain and/or pathology. This preliminary study aims to test the hypothesis that unridden horses (n = 13) have a significantly reduced frequency of occurrence of ultrasonographic changes of the SSL consistent with a diagnosis of desmitis when compared to ridden horses (n = 13) and those with clinical signs of back pain (n = 13). Results The supraspinous ligament of all horses was imaged between T(thoracic)6-T18 and ultrasonographic appearance. There was an average of 2.08 abnormal images per horse from the whole group. The average number of abnormalities in unridden horses was 4.92, in ridden horses 2.92 and in horses with clinical back pain 4.69. No lesions were found between T6 and T10 and 68% of lesions were found between T14 and T17. No significant difference (p < 0.05) was found between the three groups in the number or location of abnormal images. Conclusion The main conclusion was that every horse in this study (n = 39) had at least one site of SSL desmitis (range 2 to 11). It was clear that ultrasonographically diagnosed SSL desmitis cannot be considered as prima facie evidence of clinically significant disease and further evidence is required for a definitive diagnosis.