Abstract： Objective To establish the animal model of chronic thromboembolic pulmonary hypertension(CTEPH) and to compare the accuracy of dual-energy CT (DECT) pulmonary angiography and histopathology for detecting CTEPH. Methods Eighteen canines were included in the study. All canines underwent paracentesis, embolization, CT scanning, pressure measurement and tranexamic acid feeding. The procedures were repeated every two weeks, until systolic/diastolic pressure in canines was≥30/15 mmHg or mean pulmonary artery pressure ≥ 20 mmHg.And then canines were sacrificed for histopathology examination. For CT pulmonary angiography (CTPA)in DE mode and DECT lung perfused blood volume (Lung PBV) images, the presence or absence of PE or perfusion defects were recorded on a per-canineand aper-lobe basis. With histopathological results as reference standard, the sensitivity, specificity of CTPA and lung PBV to detect PE were computed for two readers. The pairedχ2 test (McNemar test) was used to analyze the difference in diagnostic accuracy between CTPA and Lung PBV. Inter-reader agreement was also calculated with kappa test. Results CTEPH was demonstrated in 13 canines. On a per-canine basis, both readers found uneven and peripheral perfusion defects with DECT in 11 canines (84.6%, 11/13); while 5 canines (38.5%, 5/13) had cutoff or sudden stenosis of pulmonary arteries with CTPA;on a per-lobe basis, <br> both readers had sensitivities of 14.3%(5/35), 83.3%(30/36), specificities of 100.0%(30/30), 100.0%(29/29), accuracies of 53.8%(35/65), 90.8%(59/65)for CTPA and DECT, respectively. DECT had a higher sensitivity(χ2=-4.690,P<0.01)and accuracy(χ2=8.284,P<0.01) in detecting CTEPH. Excellent and moderate inter-reader agreements were observed with CTPA and DECT (Kappa=0.938, 0.572, both P<0.001). Conclusions It is feasible to make a CTEPH animal model with autologous thrombus. DECT shows a higher accuracy than CTPA to detect CTEPH in this canine model study.