ABSTRACT. We investigated 12 anesthetized normal dogs using transesophageal echocardiography to understand the effects of respiration on the pulmonary venous flow. Additionally, we observed whether the diameter of the pulmonary vein changes with the heart beat. The pulsed Doppler wave form of pulmonary venous flow predominantly demonstrated two backward flows, with one peak occurring during ventricular systole and another during ventricular diastole. Sometimes a small forward flow occurred during left atrial contraction. In comparison with expiration, the peak velocity and velocity-time integral of the flow wave under inspiration occurred during both systole and diastole were significantly smaller. The diameter of the pulmonary vein decreased during left atrial contraction and increased during left ventricular systole and diastole.-KEY WORDS: canine, pulmonary venous diameter, pulmonary venous flow.J. Vet. Med. Sci. 61(2): 155-158, 1999 physical examination and hematology, and the ages were estimated among 1 to 3 years old according to previous report [4]. There was no evidence of cardiac or pulmonary abnormality based on either auscultation or two-dimensional echocardiography. No Dirofilaria immitis infection was identified by microfilarial detection and immunodiagnostic test (VetRED ® , AGEN Biomedical). These dogs were anesthetized for insertion of the transesophageal probe. The anesthetic protocol included premedication with atropine (0.025 mg/kg, IM) and droperidol (0.5 mg/kg, IV), induction with pentobarbital sodium (20 mg/kg, IV), and maintenance with halothane (1-1.5%) under spontaneous respiration. The dog was initially examined at left lateral recumbency. A 5 MHz biplane transesophageal probe (EUP-ES322, Hitachi Medical Corporation) connected to an ultrasound instrument (EUB-565A, Hitachi Medical Corporation) was navigated according to the method reported by Loyer and Thomas [6]. With the transesophageal probe in an unflexed position, the imaging plane was switched to transverse plane, and the probe was advanced into the stomach until the liver image was visible on the screen. The denominated middle position was achieved by slowly withdrawing and slightly flexing the probe from the stomach to a position just caudal to the point where interference by the trachea was encountered. In this position the probe was located caudal to the tracheal bifurcation and dorsal to the left atrium, between the cranial and caudal pulmonary veins.To observe the caudal pulmonary veins, we slightly advanced the probe again to a position between the right and left caudal pulmonary veins ( Fig. 1-A). This position was located between the middle and the caudal position that was described by Loyer and Thomas. The imaging plane was maintained in transverse plane to produce a long-axis four chamber view. To reduce the angle between the PVF and the Doppler sampling beam, we rotated the probe clockwise enough for adequate right caudal lobe PVF recording with pulsed Doppler mode ( Fig. 1-B). In addition,