The purpose of this study was to assess radiographic heart size in anesthetized dogs, comparing radiographs made with the patient breathing spontaneously to radiographs obtained using positive pressure manual lung inflation. The hypothesis was that manual inflation would cause reduction in the cardiac size. With dogs in right recumbency a radiograph was made at peak spontaneous inspiration followed immediately a radiograph made with application of positive pressure manual inflation of the thorax. Cardiac size was assessed, both subjectively and objectively using the vertebral heart scale (VHS). Two hundred and six dogs were studied. Manual inflation resulted in a significantly greater degree of lung inflation assessed radiographically (P < 0.0001). The subjectively assessed heart size was significantly smaller in radiographs made with manual inflation, leading to different subjective categorization of heart size in 67 (32%) of dogs. The objectively measured VHS was also significantly smaller in the radiographs made using manual inflation by a mean of 0.24 of a vertebral length (95% CL of mean difference: -0.29, -0.19, P < 0.0001). A statistically significant reduction in subjective heart size and in VHS occurred with manual inflation of the lungs, and these differences may be clinically significant. Thus, if serial radiographs are obtained to assess cardiac size, the radiographic technique used should be the same.