Aortic aneurysm and dissection account for about 2% of all deaths in industrialized countries; they are also components of several genetic diseases, including Marfan syndrome (MFS)¹. The vascular phenotype of MFS results from mutations in fibrillin-1 (FBN1), the major constituent of extracellular microfibrils^2,3. Microfibrils, either associated with or devoid of elastin, give rise to a variety of extracellular networks in elastic and non-elastic tissues³. It is believed that microfibrils regulate elastic fibre formation by guiding tropo-elastin deposition during embryogenesis and early post-natal life⁴. Hence, vascular disease in MFS is thought to result when FBN1 mutations preclude elastic fibre maturation by disrupting microfibrillar assembly. Here we report a gene-targetting experiment in mice that indicates that fibrillin-1 microfibrils are predominantly engaged in tissue homeostasis rather than elastic matrix assembly. This finding, in turn, suggests that aortic dilation is due primarily to the failure by the microfibrillar array of the adventitia to sustain physiological haemodynamic stress, and that disruption of the elastic network of the media is a secondary event.