[Purpose/significance] The paper compares industrial cluster researches from home and abroad in terms of the number of papers published, journal titles, keywords, themes, and dynamic trends [Method/process] This paper is based on the word frequency analysis, the co-word analysis and the social network analysis of papers sourced from CNKI and Web of Science (WOS) database from the time span between 1998 and 2014. [Result/conclusion] Four findings have been concluded: first, the research cycle over the past 16 years can be divided into four phases, that is, the initiation stage, the rapid growth stage, the stabilization stage and the extension and upgrading stage; second, domestic research tends to catch up with the research from abroad, and the time lag decreased from 5 years to 2 or 3 years; third, studies from overseas tend to extend while the domestic research is more likely to agglomerate; fourth, the future agenda of industrial cluster research will center on “knowledge” and “innovation” based on social network theories, while the secondary centre of research will be “industrial cluster policy (or upgrading)”, “industrial ecology and circular economy”, “firm heterogeneity”, and “cluster supply chain”.