Semantic Web and ontology have received increased attention in recent years. Various tools on modularization have been developed. However, selecting an appropriate ontology to fit a given context is difficult because assumptions for the approaches vary greatly. Therefore, a method that can compare modularization tools and help screen the most suitable approach is highly desired. In this research, a new set of evaluation criteria for selecting appropriate ontology modularization tools is proposed. Three aspects of tool evaluation, namely, tool performance, data performance, and usability are presented. This study is an empirical analysis of a number of modularization tools. Experimental results indicate that the proposed evaluation criteria for ontology modularization tools are valid and effective. It provides an evaluation method for ontology modularization, enabling ontology engineers to compare different modularization tools and easily choose the appropriate approach to produce quality ontology modules.