This paper takes 144 cities in the eastern and central China that can be largely radiated by feudal dynasty culture as the research object, and finds that the difference between rich and poor cities is closely related to the spatial distribution of ancient capitals. Based on the traceability and exploration of the historical factors about institution and culture, we establish the theoretical hypothesis of “distance from ancient capitals—marketization—urban economic growth”, that is, “distance from ancient capitals” determines the degree of the influence of feudal dynasty culture on cities outside the capital, and the feudal dynasty’s system-culture radiation which is dominated by Confucian culture decay with the increase of “distance from ancient capitals”, which leads to differences in the institutional-cultural factors that hinder the marketization of cities. Due to cultural heritage and the path dependence of institutional change, these historical institution-culture factors have potentially affected the marketization and economic growth of cities from 1978 to now.　　The empirical research uses the “distance from ancient capitals” as a historical tool variable of marketization, estimates the contribution of the marketization process to urban economic growth, and overcomes the problem of estimation bias due to endogenity. The 2SLS estimation results show that the marketization of cities has a significantly positive effect on per capita GDP. The marketization has increased by 1%, and the per capita GDP has increased by about 0.360%. When we add the control variables, or change the calculation methods of variables, empirical conclusions are very robust. Based on the research conclusions, this paper further quantitatively discusses the impact of marketization on per capita GDP: The marketization explains 62.051% of the difference of per capita GDP between Changde and Fuyang, 51.997% of the difference of per capita GDP between Changde and Dongying, and 53.139% of the difference of per capita GDP between Fuyang and Dongying. In addition, this paper uses the method of counterfactual fact which is used in the paper of Acemoglu, et al. （2015） to estimate the space for increasing the level of economic growth through accelerating marketization. If the marketization of all cities, which is less than that of Changde, would be promoted to the level of Changde, the per capita GDP of cities （144） will increase by 9.674% from the actual level. If the marketization of all cities would be promoted to the level of Dongying, the average per capita GDP of all cities will increase by 46.461%.　　This paper theoretically and empirically answers the question that “near the ancient capital” is historical wealth or historical burden, that is to say, the closer the comprehensive distance from the five ancient capitals is, the greater the blocking force of marketization in the system and culture is, which becomes the “historical burden” that affects the marketization process and economic development of these cities.