The extracellular matrix (ECM) of the aorta is a complex meshwork of elastin, collagen, and glycosaminoglycans (GAG). It also modulates the mechanical properties of the aorta, which in turn dictate lethal ruptures such as those caused by aneurysm and dissection. Amongst other roles, aortic stiffness controls the aorta’s ability to expand and recoil, and residual stresses, which are those existing in the absence of load, affect the magnitude and distribution of the mechanical stresses throughout the aortic wall. Mechanical stresses can be predicted via complex computer models, powerful tools that can also provide insight regarding the risk of rupture, given that ruptures occur when the mechanical stresses exceed the strength of the aorta. While this dissertation is primarily focused on the effect of GAG on residual stresses, other ECM (collagen, elastin) and mechanical (stiffness) factors are considered to expand our understanding of the structure-mechanics relationship in the aorta. This is important because the ECM undergoes extensive remodelling during aging and disease, but it is also critically important, as mentioned, in the context of aortic rupture. We first explored the mechanical roles of GAG in a finite element model by studying both the transmural residual stresses and the opening angle (an indicator of circumferential residual stresses) in ascending (AS) aortic ring models. Both were shown to be modulated by the GAG content, gradient, and the nature of the transmural distribution. While a heterogeneous GAG distribution led to the development of residual stresses which could be released by a radial cut, this was not the case when a homogeneous distribution was prescribed. Because the GAG distributions used in the first study were based on assumptions, and to get an in vitro understanding of the ECM role in modulating residual stresses, biomechanical mechanisms were explored in thoracic aortas from 5- to 6-month-old pigs. In a second study, we generated new detailed data on the distributions of collagen, elastin and GAG, throughout the aortic wall in the AS, arch (AR), and descending thoracic (DT) regions, and established correlations between the ECM constituents and the opening angle. The strongest correlations were observed between the opening angle and the total collagen:GAG ratio as well as the total GAG content. In line with our first in silico work, this in vitro investigation revealed that the GAG content and gradient modulate circumferential residual stresses and suggested that the interaction between GAG and the ECM fibers also plays a role in regulating residual stresses. In a third study, we examined the extent of contribution of GAG to circumferential residual stresses and to the radial compressive stiffness of the aortic wall, as well as the underlying mechanism through which GAG contribute to the mechanical properties using enzymatic GAG depletion. GAG depletion was associated with a decrease in the opening angle, by approximately 25%, 32%, 42% in the AS, AR, and lower DT regions respectively, and an increase in the radial compressive stiffness of the AS aorta. Glycation was also associated with a decrease in the opening angle, in which GAG depletion also had a similar effect. A small loss of water content was detected after GAG depletion, and the AS region was also associated with a significant loss of compressive deformation in the inner layer of the aorta following GAG depletion, suggesting that GAG interact with ECM fibers in their effect on aortic mechanics. The garnered experimental geometrical data and intramural GAG distributions were finally used to simulate animal-specific aortic rings from the AS, AR, and DT regions. The opening angle response was evaluated in solid matrices assuming one layer, and two layers to capture the different mechanical behaviors of the intima-media and the adventitia. A Holmes-Mow constitutive relationship was used and material parameters were obtained by curve fitting experimental stress-strain curves obtained from biaxial tests. Numerical results were evaluated by comparing simulated and experimental opening angles, revealing a notable overall agreement between the two.