For rectal cancer patients without nodal metastases the identification of unfavourable factors can be helpful for the better selection for adjuvant therapy and multimodality treatment. The aim of this study was to evaluate the impact of clinico-histological parameters on prognosis in node-negative rectal cancer patients. One hundred and thirty-nine consecutive node negative rectal cancer patients with complete five-year follow-up were studied prospectively. All of them underwent curative anterior resection with total mesorectal excision technique. Seventy-eight patients with tumour penetration beyond the bowel wall received neo-adjuvant short-course radiation (25 Gy) followed by surgery within 1 week and postoperative chemotherapy with 5-fluorouracil and folinic acid in six cycles or adjuvant radiochemotherapy: irradiation (50.4 Gy) combined with chemotherapy (as above). Cancer-specific survival was calculated according to the Kaplan-Meier method. Variables significant in univariate analysis by log-rank test (P < 0.05) entered the Cox proportional hazard model. Survival was decreased for males, older patients (>60 years) with extraperitoneal, poorly differentiated cancers, tumours with mucinous histology and with the absence of lymphocytic infiltration but with the lack of statistical importance. Prognosis was significantly improved for patients with T2 tumours versus T3 (P < 0.01) and with cancers with expanding growth comparing to diffusely infiltrating ones (P < 0.01). In multivariate analysis these parameters significantly and independently influenced survival (P < 0.01 and P < 0.05, respectively). Diffusely infiltrating growth of tumour can reflect the more aggressive cancer behaviour and unfavourable course of disease despite the optimised local control. Apart from the extent of tumour penetration the type of invasive margin can be an additional parameter helpful for the optimal treatment planning and better patient selection for postoperative chemotherapy.