The efficacy and safety of rigid oesophagoscopy in diagnostic and therapeutic settings in a consecutive series of 404 patients with oesophageal carcinoma were studied and compared to that for flexible oesophagoscopy in the same group. In addition, we examined the same parameters in a smaller group who had undergone radiotherapy with subsequent malignant stricturing. We performed 328 rigid procedures and 118 flexible procedures in a single regional surgical referral unit over a 7 year period. The combined perforation rate was 1.3%, with an overall mortality of 1% from 446 procedures. We conclude that rigid oesophagoscopy in the presence of carcinoma retains an important diagnostic and therapeutic role which can be achieved with low incidence of perforation in high-risk patients.