Despite improved surgical treatment strategies for rectal cancer, 5-15% of all patients will develop local recurrences. After conservative surgery, circumferential resection margin (CRM) involvement is a strong predictor of local recurrence. The consequences of a positive CRM after total mesorectal excision (TME) have not been evaluated in a large patient population. In a nationwide randomized multicenter trial comparing preoperative radiotherapy and TME versus TME alone for rectal cancer, CRM involvement was determined according to trial protocol. In this study we analyze the criteria by which the CRM needs to be assessed to predict local recurrence for nonirradiated patients (n = 656, median follow-up 35 months). CRM involvement is a strong predictor for local recurrence after TME. A margin of < or = 2 mm is associated with a local recurrence risk of 16% compared with 5.8% in patients with more mesorectal tissue surrounding the tumor (p <0.0001). In addition, patients with margins < or = 1 mm have an increased risk for distant metastases (37.6% vs 12.7%, p <0.0001) as well as shorter survival. The prognostic value of CRM involvement is independent of TNM classification. Accurate determination of CRM in rectal cancer is important for determination of local recurrence risk, which might subsequently be prevented by additional therapy. In contrast to earlier studies, we show that an increased risk is present when margins are < or = 2 mm.