This study aims to update the experience of multimodality approaches in the management of oesophageal cancer that have been adopted in several Australian and New Zealand hospitals. Between 1984 and 1985, 92 patients received pre-operative radiotherapy (30-36 Gy over 3 weeks) and one of two chemotherapy regimes (one or two courses of i.v. cisplatin 80 mg/m2 plus a 4-5 day continuous i.v. of fluorouracil 5-800 mg/m2/day) concurrently prior to surgery. Eighty-two patients (89%) underwent resection as planned. Operative specimens were microscopically free of residual tumour in 18 patients. Eight patients (9%) had treatment-related deaths: seven from surgery and one due to pre-operative chemoradiation. The Kaplan-Meier 5-year cause-specific survival estimates were 32.9 +/- 7.8% for the 58 patients with squamous cancer and 0% for the 32 with adenocarcinoma. Complete pathological response to the pre-operative regime was more common in females and was associated with a survival advantage. Five-year cause-specific survival expectation in patients who experienced a complete pathological response was 71.5 +/- 12.4%, whereas it was only 15.9 +/- 5.6% in patients who had residual cancer in their surgical specimens. Although less toxic the pre-operative regime utilizing only one cycle of chemotherapy was no less efficacious either in producing a complete pathological response or in terms of survival expectation. This uncontrolled pilot study has produced encouraging long-term results, especially for patients with squamous carcinoma that experienced a complete response to pre-operative synchronous chemoradiotherapy. A randomized controlled study comparing surgery alone with (one cycle) chemoradiation followed by surgery is now underway.