Biomedical Engineering (BME) is a rapidly growing discipline in global education due to tremendous needs for knowledge and devices to improve health care diagnosis and treatment. This paper focuses on assessing research efficiency of BME academic units. A Data Envelopment Analysis (DEA), non-parametric frontier model, is proposed to measure research efficiency of each academic units comparing to other units under consideration. Results of this approach give a single efficiency measurement for each academic unit. The results of simulated data from Thai BME academic units show that the average efficiency score is approximately 67%. This method provides university administrators a framework to evaluate research performance efficiency of nation-wide BME academic units and insights for resource allocation planning in the way to enhance their units’ efficiency.