This study was designed to evaluate whether the revised 2010 Tumour Node Metastasis (TNM) staging system could lead to a more accurate prediction of the prognosis of renal cell carcinoma (RCC) patients. A total of 1216 patients who had undergone radical nephrectomy or partial nephrectomy for RCC from 2003 to 2011 were enrolled. All of the patients had pathologically confirmed clear cell RCC (ccRCC). All cases were staged by both the 2002 and 2010 TNM staging systems after pathological review, and survival data were collected. Univariate and multivariate Cox regression models were used to evaluate cancer-specific survival (CSS) and progression-free survival (PFS) after surgery. Continuous variables, such as age and tumour diameter, were calculated as mean values and standard deviations (s.d.) or as median values. Survival was calculated by the Kaplan-Meier method, and the log-rank test assessed differences between groups. Statistically significant differences in CSS and PFS were noted among patients in T3 subgroups using the new 2010 staging system. Therefore, the revised 2010 TNM staging system can lead to a more accurate prediction of the prognosis of ccRCC patients. However, when using the revised 2010 staging system, we found that more than 92% of patients (288/313) with T3 tumours were staged in the T3a subgroup, and their survival data were not significantly different from those of patients with T2b tumours. In addition, T2 subclassification failed to independently predict survival in RCC patients.