This paper utilizes cointegration theory, error correcting model and Granger causality testing theory to make an empirical research on the relation between urbanization and GDP in China, and also implements a comparative analysis to the relation between three industries and degree of urbanization, the related coeffecient is 0.97, 0.95, 0.97, 0.97. And the result shows a long-term balance between these two factors, and the promoting effect to tertiary industry by urbanization is more obvious. Urbanization and economic growth are the long-term balanced relations. In the long-term balance, every 1% increment of urbanization can make 4.82% increment of GDP; In short-term balance, if the balance depart from the long-term balance at the i-th term, the model will take automatic reversal adjustment with -0.06 adjusting strength at the (i+1)th term, to make it move to the long-term balance. The economic growth onto urbanization is one-way causality relationship, the primary and secondary industry onto urbanization is also one-way causality relationship. However, the tertiary industry onto urbanization is both-way causality relationship.