Aims Angiotensin II-infused ApoE(-/-) mice are a popular mouse model for preclinical aneurysm research. Here, we provide insight in the often-reported but seldom-explained variability in shape of dissecting aneurysms in these mice. Methods and results N=45 excised aortas were scanned ex vivo with phase-contrast X-ray tomographic microscopy. Micro-ruptures were detected near the ostium of celiac and mesenteric arteries in 8/11 mice that were sacrificed after 3 days of angiotensin II-infusion. At later time points (after 10, 18, and 28 days) the variability in shape of thoraco-abdominal lesions (occurring in 31/34 mice) was classified into 7 different categories based on the presence or absence of a medial tear (31/31), an intramural hematoma (23/31) or a false channel (11/23). Medial tears were detected both in the thoracic and the abdominal aorta and were most prevalent at the left and ventral aspects of celiac and mesenteric arteries. The axial length of the hematoma strongly correlated to the total number of ruptured branch ostia (r(2) = 0.78) and in 22/23 mice with a hematoma the ostium of the left suprarenal artery had ruptured. Supraceliac diameters at baseline were significantly lower for mice that did not develop an intramural hematoma, and the formation of a false channel within that intramural hematoma depended on the location, rather than the length, of the medial tear. Conclusion Based on our observations we propose an elaborate hypothesis that explains how aortic side branches (i) affect the initiation and propagation of medial tears and the subsequent adventitial dissection and (ii) affect the variability in shape of dissecting aneurysms. This hypothesis was partially validated through the live visualization of a dissecting aneurysm that formed during micro-CT imaging, and led us to the conclusion that angiotensin II-infused mice are more clinically relevant for the study of aortic dissections than for the study of abdominal aortic aneurysms.