Introduction: Barrett’s oesophagus (BO) is a metaplastic premalignant disease which can undergo a metaplasia-­‐dysplasia-­‐adenocarcinoma pathway. It represents an example of field cancerization by which an area occupied by BO can undergo molecular and genetic changes associated with carcinogenesis without being phenotypically cancerous. Previous work suggested that non-­‐cancerous BO contains a monoclonal population. More recent work demonstrated that premalignant Barrett’s fields are polyclonal suggesting that clonal interactions may be important in carcinogenesis. It is the aim of this thesis to further investigate clonal interactions in BO by understanding the effects of therapy in altering the relationships of clonal populations in BO, by assessing the relationship of clonal populations in dysplasia as compared with the associated cancer, and by attempting to elucidate a potential molecular mechanism of clonal interactions. Results: The overall results can be summarised as follows: 1.Premalignant clonal populations are well mixed allowing for clonal interactions. However, the adenocarcinoma associated with high grade dysplasia is monoclonal and derived from clonal populations found in the dysplasia, indicating possible clonal interactions during carcinogenesis. 2. Patients with persistent disease after endoscopy retain the same clonal populations. However, the clonal populations of recurrent disease changes such that new clonal populations arise or may benefit from the extinction of others. 3. These clonal populations may be derived from deep submucosal glands or may be found in phenotypically normal squamous epithelium indicating a common stem cell origin. 4. A possible mechanism of clonal interaction may be the senescence associated secretory phenotype: senescence is abundant in BO and can cause proliferation in neighbouring cells in vitro. Conclusion: This thesis has investigated the implications of clonal interactions in BO. The demonstration of temporal clonal heterogeneity as a result of endoscopic therapy, as well as spatial clonal heterogeneity possibly resulting in carcinogenesis, asks for a mechanistic explanation of clonal interactions. The consequences of senescence may well provide one such mechanism.