The current TNM (tumor, nodes, metastases) staging system for human hepatocellular carcinoma (HCC) has been challenged since a new T staging system was proposed to correlate the staging group with patient outcome after curative liver resection. The new T staging system proposed T1 as no vascular invasion, small size (< or = 5 cm), and solitary tumor. T2 was defined as the presence of one of the following factors: size greater than 5 cm, vascular invasion, or multiple tumors; T3 as the presence of two of the above three factors; and T4, the presence of all three factors. A total of 323 patients undergoing curative partial hepatectomy for HCC were studied. Kaplan-Meier survival analysis was used to evaluate the postoperative outcome. The new T staging showed good correlation between the staging group and patient outcome. The 1-year disease-free survival (DFS) rate and overall survival (OS) rate were 80.0% and 87.8% for stage 1 (n = 115), 67.6% and 81.6% for stage 2 (n = 136), 40.0% and 58.0% for stage 3 (n = 58), and 21.4% and 42.8% for stage 4 (n = 14), respectively. The 3-year DFS rate and OS rate were 61.0% and 64.5% for stage 1, 37.8% and 50.7% for stage 2, 21.4% and 29.8% for stage 3, and 21.4% and 34.3% for stage 4, respectively. When analyzed using the current International Union Against Cancer (UICC) pathologic (p) TNM staging system, the 1-year and 3-year DFS rates were 86.2% and 64.0% for stage 1 (n = 30), 73.9% and 50.0% for stage 2 (n = 182), and 46.8% and 22.3% for stage 3 (n = 111), respectively. Our results showed that, while both staging systems allow clear stratification of patients into prognostic groups, the modified TNM system is not superior to the UICCpTNM system in predicting survival of HCC patients after curative partial hepatectomy. A larger scale, multicenter study may be needed to test the revised TNM system.