In this paper, I use panel data to investigate the relationship between bank performance and non-interest income. The data contains annual observations of 15 representative Chinese commercial banks for the period of 2007-2015. Based on likelihood ratio test and Hausman test, the fixed effect model for panel data has been used to analyze the result. Our empirical results indicate that non interest income can improve financial performance of Chinese commercial banks measured by Return on Assets (ROA) with some degree. However, It has a negative influence on profit of state-owned banks. Net interest spread and cost-to- income rate still play the dominant role in the bank performance. Besides, the empirical result shows that there is a negative correlation between annual GDP growth rate and bank performance, which seems to contradict with general economic theory. The possible reason is that the most of measures to stimulate the economy taken by government in China due to the financial crisis during the selected sample period were achieved by bank credit. Meanwhile, the lending rate was relatively high at that time. Accordingly, the sufficient liquidity and elevated spreads made better performance of banks though under the background of international finance crisis and economic downturn