With the advances in research of the Tibetan Plateau (TP) and the deep understanding of multi-disciplinary research, coupled with the progress of geographic big-data, earth observation science and technology, this research systematically discussed the principles, methods and basis for determining the boundaries of the TP. It analyzed the macro landform structures (plateau surface, low basin and deep-cut valley lowland on the edge of the TP) within the TP and the fundamental characteristics of the geographic units' composition in TP's surrounding areas. Based on the high resolution remote sensing images, DEM data and geomorphologic maps etc., the boundary of TP with a 1:1000000 scale is defined through a comparative study of geomorphological features with the support of Arcmap 10.5. The results show that the Tibetan Plateau stretches from the foot of the Himalayas in the south to the foot of the Kunlun Mountains and the Qilian Mountains in the north, with a total length of 1560 km. While it spans about 3360 km from the Hindu Kush Mountains and the Pamir Plateau in the west, to the eastern foot of Hengduan Mountains in the east. The TP, lying between 25°59′30″N-40°1′0″N and 67°40′37″E-104°40′57″E, covers a total area of 3083.44 × 103 km2, with an average altitude of 4320 m. Geographically, the TP is located in Southwest China and eight other countries including India, Pakistan, Tajikistan, Afghanistan, Nepal, Bhutan, Myanmar and Kyrgyzstan. The TP in Chinese section has an area of 2580.90 × 10 3 km2, accounting for around 83.7% of the total area, with an average altitude of 4400 m. In China's part, the TP spans in six provincial-level regions: Tibet Autonomous Region (TAR), Qinghai province, Gansu province, Sichuan province, Yunnan province and Xinjiang Uygur Autonomous Region. Among them, the main parts of TAR and Qinghai are the major section of the TP, which accounted for 60.6% of the total area of the plateau.

Keywords： Tibetan Plateau; principles for determining the boundary; method basis; boundary; plateau characteristic data