Initiatives on developing regional economic clusters in modern conditions have great importance. To build efficient cluster policy in Russian regions it is necessary to renovate methodical approaches to processes of structural identification and classification of clusters. The paper highlights problems of structural identification and classification of regional economic clusters. Systemic features of clusters are distinguished from traditional integrated business structures that implement strategies for diversification and vertical integration. The approaches to the structural identification and classification of regional economic clusters, both in terms of industry affiliation, are included in their composition of enterprises, and the structure and scale of economic activity are systematized. Particular attention is paid to the application of the theory of fuzzy sets for the cluster identification. Within the framework of this approach, six basic features of the cluster are integrated in the formal membership function: the competition between the cluster members, territorial localization of the main elements of the cluster, entering the world markets, branch belonging, use of the common infrastructure of the region, presence of integration interaction. The process of identification of cluster structures in the metallurgical complex of the Volgograd region on the basis of a matrix composition was carried out. The applicability of the theory of fuzzy sets for different types of regional clusters, according to M. Enrait's classification is substantiated: workers, latent, potential, and also by structural classification: "frame" and "diffuse" clusters. Recommendations are given to bodies of state power and administration on the improvement of regional cluster policy.