PURPOSE: The aim of our study was to identify risk factors associated with anastomotic leakage (AL) after an anterior resection (high anterior resection+low anterior resection) for rectal cancer. METHODS: Between January 1998 and December 2007, 356 patients underwent an anterior resection for rectal cancer. Early anastomotic leakage (EAL) was defined as leakage identified during hospitalization. Late anastomotic leakage (LAL) was defined as leakage identified in outpatients. RESULTS: AL (EAL+LAL) occurred in 30 patients (8.4%, mean time: 15.4 days). Among of them, EAL occurred in 20 patients (5.6%, mean time: 5.1 days), and LAL occurred in 10 patients (2.8%, mean time: 36.0 days). In the univariate analysis, the size of the tumor, the tumor level from the anal verge, and the level of anastomosis were significantly associated with AL. In EAL, the size of the tumor, the tumor level from the anal verge, the level of anastomosis, the operation type, and the value of serum albumin on day 3 after the operation were risk factors. In LAL, the tumor level from the anal verge and the level of anastomosis were risk factors. In the multivariate analysis, tumor size >7 cm (AL: P<0.001, EAL: P<0.001) and tumor level from the anal verge < or =8 cm (AL: P=0.014, EAL: P=0.001) were independent risk factors. CONCLUSION: AL and EAL after an anterior resection for rectal cancer were related to the size of the tumor and the level of the tumor from the anal verge.