Abstract: The evolution of urban landscape pattern and spatial form is a response to rapid urbanization, and the special geographical environment of  the areas surrounding lake in western plateau makes the spatial and temporal evolution of urban settlements different from plains, mountains and other regions. Landsat TM/ETM+/OLI remote sensing imageries were used to interpret urban patches in Dianchi region from 1990 to 2018, and landscape indices were calculated and then used to deeply analyze the evolution characteristics and morphological regularities of sixteen towns. The results show that from 1990 to 2018, urban settlements in the areas experienced the undulated evolution which displays a “low-speed growth-high-speed growth-rational growth” pattern. From 2000 to 2010, urban settlements expanded fastest and the expansion rate had reached 54.2%. The patterns of evolution in different periods and different directions were significantly different. The spatial characteristics and development direction of town patches are obviously non-independent and compact. Using K-means clustering to divid urban patches into four types of agglomeration, the result showed that the large patch dominant type accounted for 6.7%, high-density cluster type accounted for 35.8%, medium-density cluster type and low-density scattering point type accounted for 48.6% and 8.9%, respectively. In term of the spatial pattern, urban expansion of these areas appeared as a “central-peripheral” expansion characteristic: Qianwei, Yiliu, and other peripheral neighborhoods expanded rapidly, while Biji, Haikou, and other central neighborhoods grew slowly. In Wuhua and Guandu regions, the numbers, degrees of aggregation and cohesion of urban patches showed an increasing trend, and both the connectivity and aggregation between towns were increasing with the increase of patch numbers, though these two indices had little impacts on fragmentation of patches. The town patches were mainly distributed in areas with an altitude of 1 900-2 100 m and a slope of less than 9°. They had strong correlation with population and GDP, and showed strong transportation and river directionality.