Aims: To determine the prognostic significance of the nodal stage and number of nodes recovered in the surgical specimen after preoperative synchronous chemoradiation (SCRT) and surgery for locally advanced or unresectable rectal cancer.Materials and Methods: One hundred and eighty-two consecutive patients with locally advanced or unresectable (T3/T4) rectal carcinomas were entered on a prospective database and treated in this department with preoperative chemoradiation, followed 6-12 weeks later by surgical resection. Most patients received chemotherapy in the form of low-dose folinic acid and 5-fluorouracil (5-FU) 350 mg/m(2) via a 60-min infusion on days 1-5 and 29-33 of a course of pelvic radiotherapy delivered at a dose of 45 Gy in 25 fractions over 33 days to a planned volume. After resection, patients with a positive circumferential margin (<= 1 mm), extranodal deposits or Dukes' C histology received adjuvant 5-FU-based-chemotherapy (n = 40).Results: After SCRT, 161 patients underwent resection. Twenty-one patients remained unresectable or refused an exenterative operation. Median follow-up is 36 months. Down-staging was achieved in most patients, with 19 having a complete pathological response (pT0). The median number of lymph nodes recovered for all patients was five (range 0-21). The 3-year survival rate for node-positive patients is 47%, for node-negative patients with less than three lymph nodes recovered is 62% and for node-negative patients with three or more lymph nodes recovered is 70%. Compared with node-positive patients, simple regression models revealed a reduced hazard ratio (HR) of 0.72 (0.36-1.43) for node-negative patients with less than three nodes recovered and 0.48 (0.26-0.89) for node-negative patients with three or more lymph nodes recovered. In a multivariate model, including nodal status, excision status, age and sex only positive excision margins significantly predicted a poor outcome: HR = 3.05 (1.55-5.97).Conclusions: The number of nodes found after preoperative chemoradiation is a significant prognostic factor by univariate analysis. In this study, patients with node-negative histology, and at least three nodes recovered, had better long-term survival than patients in whom two or less nodes were recovered or with positive nodes. This effect was attenuated by the inclusion of excision status in multivariate models.