Refactoring has become an essential part of software development process especially for large and long lasting projects. Extract-class is one of the vital refactorings that is used to improve cohesion of a class by splitting large in-cohesive classes into more cohesive ones. Providing automated means of identifying opportunities for extract-class refactoring could make the software maintenance efficient. In this paper, a novel algorithm viz. ESA ("Exclusive-Similarity" Algorithm) is proposed to identify extract-class refactoring candidates automatically. The algorithm proposes new metrics viz. Exclusive-Similarity metric (ESM), Cohesion among Method-Clusters, (CMC) and Method-Similarity with Attribute-Clusters (MSAC). The proposed algorithm has been realized into a tool as an add-in to Visual Studio and the tool is exercised with 3 open-source projects to demonstrate applicability.