Abstract: Taking Yining city, Xinjiang province as an example, based on Maximum likelihood classifie（MLC） method with remote sensing images of 2000,2005,2010,2015 and 2018, the authors classified the land of Yining and extracted the urban proper to analyze the urban construction land expansion and spatial distribution in the past 19 years. Urban expansion indices, which included expansion speed index（ESI）,expansion intensity index（EII）, expansion degree index（EDI） and standard deviational ellipse（SDE）, were used to quantitatively analyze temporal variations of construction expansion. The following results were obtained：①The construction land area of Yining city increased by 80.03 km² and it was the 2.7 times of the initial urban from 2000 to 2018, According to the analysis of SDE, the construction land extended to the northwest and the urbanization showed a rapid expansion trend from 2000 to 2005. ②During the period of 2000—2018, the construction land expansion of Yining city was mainly realized through occupying large areas of cultivated. In the growth of the construction of the city area, city expansion rate increase year by year, city expansion strongly expands and then the expansion tends to slow down. ③The urban expansion of Yining city was affected by both natural and human factors. Natural factors affected the space scope of urban expansion, and human factors was the direct driving force for construction land changa in city expansion.

Key words: supervised classification, urban spatial expansion, transition matric, Yining city