The European Organisation for Research and Treatment of Cancer (EORTC) 22921 four-arm randomised trial questioned the value of preoperative chemoradiation (XRT-CT) versus preoperative radiation (XRT) and the value of additional postoperative chemotherapy (CT) versus none in T3-T4 M0 resectable rectal cancer patients. We report on the preoperative toxicity, treatment compliance and early deaths (all deaths up to 30 days after surgery) of the two treatment modalities in patients who were entered into trial before January 2001. In the XRT Group (group A), patients received 45 Gy, 25 fractions over 5 weeks. In the XRT-CT Group (group B), two 5-day courses of CT were added to the first and fifth weeks of XRT. For each CT course: 5-fluorouracil (5-FU) 350 mg/m2/day and Leucovorin (LV) 20 mg/m2/day were given. 398 and 400 patients started treatment in groups A and B, respectively. Grade 2+acute diarrhoea occurred in 17.3 and 34.3% of patients in groups A and B, respectively (P<0.005). The other side-effects remained unchanged or were only marginally increased. The compliance with RT was 98.5 and 95.5% in groups A and B, respectively. In group B, 78.7 and 71.1% of the patients received 95-105% of the planned CT doses at the first and second courses, respectively. 6 patients died preoperatively, 2 from toxicity in group B. 8 patients (1%) died within the 30 days after surgery in both groups. At the doses recommended in the protocol, the addition of 5-FU-LV to preoperative XRT slightly increased the amount of acute toxicity. However, the compliance with the radiation protocol or the feasibility of surgery did not decrease.