When knowledge is a paramount resource, the effective management of its accumulation, use, and transfer becomes the most important function, which enables to assess and monitor the ability of territorial innovation systems to generate new scientific knowledge. While the commercialization of knowledge in the territorial context is primarily considered at the regional level as part of the production process, it is more appropriate to consider the knowledge generation at the level of a city where the city is the most important urbanized scientific regional centre. This study aims to assess the territorial distribution of Russian cities by their ability to generate new scientific knowledge with the subsequent development of recommendations for managing their territorial capital as the basis of competitiveness. For Russia, the solution to this problem is most relevant in the view of a wide variety of cities and the impossibility and even danger of using unified approaches to their development. The large territory of the country, the remoteness of individual cities from each other, their diversity in size, development level, knowledge specialization makes it important to manage the diversification of knowledge. In this study, a science-related approach was used to assess the knowledge generation based on the publication activity analysis of 2013—2017 according to the international abstract database of Scopus scientific citation. The study includes 440 cities with at least 1 publication. Based on the results of the study, Russian cities fall into three types and individual subtypes, considering their specific features. The first type of cities “above average” is specified with high indicators of academic productivity, while for the first subtype they are supplemented by high indicators of scientific connectivity, citation and quality of publications. For the second subtype, these indicators are significantly lower. The second type of cities, “medium” or “transitional” have significant publication experience. However, the insufficient average citation level and very few publications in high-ranking journals do not allow them to be attributed to the leading national scientific centres of the first group. The third type “below average” with the lowest values of all indicators is divided into three subtypes: the first group includes cities whose scientific groups publish a small number of articles with a high level; the second group brings together the cities with a low level of scientific productivity having scientific connections at the national and international level; the third, the “silent” cities, which do not show any activity in any way.