Ignoring the realistic basis and blindly pursuing new industries have become a common phenomenon in Chinese cities. Scholars have generally called for the implementation of regional capacity-based development policies, taking into account regional realities and industrial base, to avoid "one size fits all", repeated construction and vicious competition. Based on the patent data of the State Intellectual Property Office (SIPO), this paper describes the spatial-temporal evolution characteristics of technology relatedness and complexity of Chinese cities from 1987 to 2016, and estimates the impact of technology relatedness and complexity on regional technology introduction and technology growth. The results show that: (1) The dynamic of technology entry and exit in the eastern region is stronger than that of the central and western regions. The correlation analysis between relatedness and technology entry and exit shows that technology with high relatedness is more likely to enter the region. In contrast, technologies are more likely to exit if they do not have strong local technology linkages. (2) From the view of technological relatedness and complexity, technological evolution in eastern coastal areas is more dependent on the technological relatedness, and technological complexity has experienced tremendous growth, while technological relatedness and complexity in the central and western regions have changed little, and technological development in Chinese cities generally presents path-dependent characteristics. (3) Technological relatedness has been playing a significant role in promoting the entry and growth of new technologies. When technology with high complexity is introduced, if it has a high relatedness with the existing technological foundation of the region, it will play a significant role in promoting regional technological progress; on the contrary, even if the blind introduction of complex technology will not have a significant impact on regional technological progress. Based on the relatedness and complexity, this paper divides the technology development paths into four types: The policy of "bright road" means low risk and high benefit, when technologies are expected to exceed average returns under relatively low risk. In contrast, technological development with high risk and low benefit are unlikely to catch and raise value, we therefore refer to such a policy as "dead end". The "road exploration" policy means high risk and high benefit, which aims to develop new and original technology to realize path creation. "Slow Road" policy means driving safely forward in the existing low-level technology path. The framework of regional diversification development can provide a reference for decision makers to weigh and choose the path of technological development.