Aims: The aim of this review is to consolidate our knowledge on an important and rapidly expanding area of expertise. Numerous methods for predicting response (in terms of pathological response and survival) to neoadjuvant therapy (chemotherapy/chemo-radiotherapy) in oesophageal and junctional cancers have been proposed. This review concerns itself only with the use of positron emission tomography for such a purpose. At present there are no standardised criteria amongst PET trials as to what determines a response according to PET, what is the optimal time to perform PET in relation to the timing of neoadjuvant therapy, and what is the ideal method of quantifying PET tracer uptake. Methods: An electronic search was performed of PubMed, Ovid and Embase websites to identify studies, in the English language, using the search terms: PET; oesophageal; oesophago-gastric; survival; cancer; response; chemotherapy and chemo-radiotherapy. The reference lists were searched manually to identify further relevant studies. Results: Twenty-two studies were identified, all using (18)FDG as the tracer, using PET to predict response in terms of pathological response and survival following neoadjuvant therapy (chemotherapy/chemo- radiotherapy). PET had a varying degree of success in predicting both pathological response and survival outcomes, with only one study using PET to influence management decisions. Conclusions: PET seems a promising technique, but large-scale conclusions are hindered by small study numbers, lack of criteria as to what constitutes a response and markedly differing PET imaging times. A large randomised trial concerning a homogeneous group of patients and tumours is required before PET might be used to influence management. (C) 2009 Elsevier Ltd. All rights reserved.