To realize process-driven management, so many companies have been launching business process management systems. Business process is collection of standardized and structured tasks inducing value creation of a company. Moreover, it is recognized as one of significant intangible business assets to achieve competitive advantages. This research introduces a novel approach of workflow process analysis, which has more and more significance as process-aware information systems are spreading widely into a lot of companies. In this paper, a methodology of workflow clustering based on process similarity has been proposed. The purpose of workflow clustering is to analyze accumulated process definitions in order to assist design of new processes and improvement of existing ones. The proposed methodology exploits measures of structural similarity of workflow processes. The methodology has been experimented with synthetic process models for illustrating the implication of workflow clustering.