This thesis comprises two separate molecular studies on Hereditary Haemorrhagic Telangiectasia (HHT) families with pulmonary arteriovenous malformations (PAVMs). PAVMs occur in over 25% of patients with the autosomal dominant disorder HHT. They account for some of the most devastating consequences of the disease. Mutations in two genes, endoglin and ALK1, are known to cause HHT. Both encode a protein expressed on vascular epithelial cells and are involved in signalling by members of the transforming growth factor (TGF)β superfamily. To date, PAVMs have not been detected in ALK1 families. It has been suggested therefore that clinical screening for PAVMs is restricted to endoglin-linked families. There is evidence from a single HHT family without pulmonary involvement that a third HHT gene may exist. In one study in this thesis the endoglin promoter region was screened for mutations in four endoglin-linked families, and four other small families, in whom no mutations had previously been found studying all coding regions and splice sites of the gene. No mutations were found. However a two base pair CA deletion polymorphism, with an allelic frequency of 9% in the UK population, was found 2563 base pairs upstream of one of the major endoglin transcription start sites. It may prove a useful tool for association studies. In the other study linkage analyses were performed on four PAVM-HHT families, two of which were expanded during this study. They had all previously been found to be unlinked to endoglin. Linkage to both endoglin and ALK1 was significantly excluded in one family. Significant linkage to ALK1 was found in another. It is concluded that a third HHT gene exists and that PAVMs can occur in all genotypes with resulting clinical implications. Confirmation by this work that at least a third HHT gene exists should precipitate identification of the gene and elucidation of its biochemical role. It seems reasonable to speculate that it will encode another component of the TGFβ signalling complex or a downstream effector, and so its identification should increase our understanding of this superfamily of signalling molecules. Further work however is still required to determine how the actions of the TGFβ signalling complex lead to abnormal vascular structures.