This paper deals with the concept of differentiated knowledge bases, an area related to the study of the regional innovation systems. The concept explores the ways in which knowledge is involved in innovation processes and knowledge flows within and between organisations. Some limitation of this concept can be the fact that there is no clear method for identification of knowledge bases in regions. The aim of this paper is to propose appropriate indicators and methodology to empirically identify the knowledge base of a region. The second aim of this paper is to use this methodology for the identification of knowledge bases in the Czech regions. The methods used for the classification include the point method and cluster analysis. Based on the combination of these methods, the Czech regions are classified into four groupsthose with a prevailing analytical, synthetic, analytical-synthetic, and synthetic-analytical knowledge base. The Capital City of Prague, the South Bohemian Region, and the South Moravian Region were among the ones with an analytical base. A synthetic knowledge base was identified in the Liberec Region, the Zlin Region, and the Vysocina Region. One of the two transitional types was found in the remaining regions. In real-life practice the findings can be applied, above all, to the implementation of regional innovation policies, which should always respect the characteristics of the given region.