In order to assess the influence of the vertebral heart scale (VHS) on the accuracy of the radiographic diagnosis of cardiac disease, thoracic radiographs of 50 dogs with proven cardiac disease, 26 with other thoracic diseases, and 50 with no clinical signs of cardiovascular or respiratory disease were mixed and examined by three independent, blinded observers chosen to represent a range of radiographic abilities. They first examined all the radiographs without making measurements of VHS and made a diagnosis. They then re-examined the radiographs, and measured VHS on both lateral and dorsoventral or ventrodorsal radiographs before again recording a diagnosis without reference to their original diagnoses. For all the observers, the dogs with cardiac disease had a higher mean VHS than the normal dogs. A VHS over 10.7 on the lateral radiograph was a moderately accurate sign of cardiac disease. The observers' accuracy of diagnosis did not change significantly as a result of using VHS as an adjunct to a subjective assessment of the radiographs.