Aim Recent Medical Research Council trials in UK have shown significantly improved disease free survival for patients receiving preoperative chemotherapy for oesophageal and gastric cancer, albeit only by a few months. The aim of the study was to study in isolation other factors affecting survival in consecutive patients undergoing surgery irrespective whether they were recruited to participate in the trial. Methods Between April 1999 and Dec 2003 a total of 56 patients diagnosed to have operable gastric (n=33) and oesophageal adenocarcinoma (n=23) after staging (endoscopy, thin slice spiral CT, endoluminal ultrasonography, staging laparoscopy) underwent radical resection with extended lymphadenectomy. 29 patients were randomised to receive neoadjuvant chemotherapy (3 cycles of epirubicin, cisplatin, 5FU pre- and post-operative for gastric cancer and 2 cycles of cisplatin, 5FU preoperative for oesophageal cancers. Follow-up was based on clinical progress and investigations only in suspected recurrence or patients with progressive weight loss. Results There was 0% 30 day post-operative mortality. One patient died on day 48 from cachexia due to advanced disease. The mean survival for chemotherapy + surgery (CS) group was significantly longer 27 months versus 17 months for surgery (S) alone group(p=0.005). There was no difference in the overall survival. The most important prognostic factor in the survival outcome was the T staging. 14/21 patients with early stage (T1/2 N0) cancer were long term survivors and disease free at median follow-up of 3 years. 28 patients presented with more advanced stage (T3N0/1) with only nine surviving and there was no difference between the 2 groups. 7 patients were understaged and had pathological T4 lesions resulting in poor outcome(Median survival 6 months). 9/29 patients tumour were deemed downsized by the chemotherapy but this did not translate into improved survival. Conclusion Neoadjuvant chemotherapy improves progressive disease free survival but this benefit is offset by the prolonged duration of chemotherapy administration and slower recovery from surgery. Accrual of patients for trials has been based on preoperative staging which has its inherent problems with accuracy. The authors believe that pre-operative chemotherapy may improve survival outcome in patients with low volume tumours.