Purpose -This paper aims to improve the Commission of Higher Education (CHE)'s current university classification and develop the Thai Higher Education Classification Model (THEC-model). This study supports the CHE's initiative to ensure that the future is more science-oriented by encouraging universities to become National Research Universities (NRUs). Design/methodology/approach -The research applies empirical data and a statistical approach for the THEC-model's development. The model's results are then compared with the decisions reached earlier by the CHE in selecting public universities as research-intensive. Findings -The proposed classification criteria for NRUs consist of: research funding; the variety of instructional programmes; the level of instructional programmes; instructors and research staff body; and student body, which have significantly statistically influenced the differences in Y-variables: research output, citation, and research awards at alpha 0.05. The initial results show that eight universities are selected. The findings are consistent with the 2008 announcement, except for two universities. Practical implications -The developed THEC-model benefits academic researchers, university administrators, and policymakers for many reasons. For example, the THEC-model provides information for academic researchers to determine the important variables for a research university. The model provides information for policymakers to manage higher education effectively to raise the targets for a university. Originality/value -The THEC-model criteria were generated by reviewing the classification system in different locations. Such criteria could be applied extensively at domestic and international level. Moreover, the developed THEC-model is based on a statistical approach and empirical data improved the reliability and would be beneficial to the CHEs in Thailand for further improvement on research-focused HEI classification criteria in the future.