124 Journal of Applied Business and Economics Vol. 21(3) 2019 Comprehensive Efficiency Assessment of Turkish Teaching Hospitals: Technical, Pure Technical and Scale Efficiencies with Data Envelopment Analysis Cuma Yildirim New Jersey City University Hakan Kacak Ministry of Health of Turkey Selami Yildirim Azerbaijan State Economy University Sahin Kavuncubasi Baskent University The purpose of this research is to point out detail performance analysis of general teaching hospitals and investigate of the efficiency pattern of them. For this research, Data envelopment analysis (DEA) approach is used to evaluate the relative technical and scale efficiencies of general teaching hospitals. Clinical service quality development strategies must be developed to decrease hospital mortality. Hospitals must put a lot of numbers of the beds and nurses to serve more suitable scale sizes. INTRODUCTION Resource utilization and hospital efficiency have become significant issues of the health policy because of the steady growth in hospital expenses. With the most complex and expensive treatment practices, teaching hospitals are one of the leading actors in the healthcare delivery system in Turkey. The objectives of this study are comprehensive performance analysis of general teaching hospitals and examination of the efficiency pattern of them. The scope of this study is 48 Ministry of Health (MoH) teaching hospitals. Data envelopment analysis (DEA) approach is used to evaluate the relative technical and scale efficiencies of the teaching hospitals. Empirical findings show that 15 hospitals (31%) are technically efficient with the average score of 0,878. 9 hospitals (19%) are pure technical efficient but scale inefficient, 11 hospitals (23%) are both pure technical and scale inefficient, and 13 hospitals (27%) are scale efficient but pure technical inefficient. 58% of the hospitals manage on Most Productive Scale Size, 29% of them have diseconomies of scale or decreasing return to scale, and 13% of them have economies of scale or increasing return to scale. The slack analysis shows that ex-cases must be