EORTC trial 22921 examined the addition of preoperative or postoperative chemotherapy to preoperative radiotherapy in patients with rectal cancer. After a median follow-up of 5 years, chemotherapy-irrespective of timing-significantly improved local control. Adjuvant chemotherapy did not improve survival, but the Kaplan-Meier curves diverged, suggesting possible delayed benefit. Here, we report the updated long-term results. We randomly assigned patients with clinical stage T3 or T4 resectable rectal cancer to receive preoperative radiotherapy with or without concomitant chemotherapy before surgery followed by either adjuvant chemotherapy or surveillance. Randomisation was done using minimisation with factors of institution, sex, T stage, and distance from the tumour to the anal verge. Study coordinators, clinicians, and patients were aware of assignment. Radiotherapy consisted of 45 Gy to the posterior pelvis in 25 fractions of 1·8 Gy over 5 weeks. Each course of chemotherapy consisted of fluorouracil (350 mg/m(2) per day intravenous bolus) and folinic acid (leucovorin; 20 mg/m(2) per day intravenous bolus). For preoperative chemotherapy, two courses were given (during weeks 1 and 5 of radiotherapy). Adjuvant chemotherapy was given in four cycles, every 3 weeks. The primary endpoint was overall survival. This analysis was done by intention to treat. The trial is registered with ClinicalTrials.gov, number NCT00002523. 1011 patients were randomly assigned to treatment between April, 1993, and March, 2003 (252 to preoperative radiotherapy and 253 to each of the other three groups). After a median follow-up of 10·4 years (IQR 7·8-13·1), 10-year overall survival was 49·4% (95% CI 44·6-54·1) for the preoperative radiotherapy group and 50·7% (45·9-55·2) for the preoperative radiotherapy and chemotherapy group (HR 0·99, 95% CI 0·83-1·18; p=0·91). 10-year overall survival was 51·8% (95% CI 47·0-56·4) for the adjuvant chemotherapy group and 48·4% (43·6-53·0) for the surveillance group (HR 0·91, 95% CI 0·77-1·09, p=0·32). 10-year disease-free survival was 44·2% (95% CI 39·5-48·8) for the preoperative radiotherapy group and 46·4% (41·7-50·9) for the preoperative radiotherapy and chemotherapy group (HR 0·93, 95% CI 0·79-1·10; p=0·38). 10-year disease-free survival was 47·0% (95% CI 42·2-51·6) for the adjuvant chemotherapy group and 43·7% (39·1-48·2) for the surveillance group (HR 0·91, 95% CI 0·77-1·08, p=0·29). At 10 years, cumulative incidence of local relapse was 22·4% (95% CI 17·1-27·6) with radiotherapy alone, 11·8% (7·8-15·8) with neoadjuvant radiotherapy and chemotherapy, 14·5% (10·1-18·9) with radiotherapy and adjuvant chemotherapy and 11·7% (7·7-15·6) with both adjuvant and neoadjuvant chemotherapy (p=0·0017). There was no difference in cumulative incidence of distant metastases (p=0·52). The frequency of long-term side-effects did not differ between the four groups (p=0·22). Adjuvant fluorouracil-based chemotherapy after preoperative radiotherapy (with or without chemotherapy) does not affect disease-free survival or overall survival. Our trial does not support the current practice of adjuvant chemotherapy after preoperative radiotherapy with or without chemotherapy. New treatment strategies incorporating neoadjuvant chemotherapy are required. EORTC, US National Cancer Institute, Programme Hospitalier de Recherche Clinique, Ligue contre le Cancer Comité du Doubs.