Abstract: The objective of this study was to reveal the spatial distribution characteristics and change trend of cultivated land, taking the Manas River as the study area, by extracting the cultivated land information based on 1975, 1990, 1995, 2000, 2005, 2010, and 2015 series of land use data, via spatial autocorrelation, gravity center mode, standard deviational ellipse, and geography informatic spectrum analysis. The results revealed that there was significant global positive correlation and strong local spatial heterogeneity in the spatial distribution of cultivated land. The cultivated land showed an aggregated time series trend, which first weakened, then strengthened, and then weakened again across the study period. The land formed a high-high agglomerate distribution situation with a left core in Shanghudi Township of Shawan County, and a right core in Beiquan Town of Shihezi City. The overall spatial change in the cultivated land showed a clear, significant shift to the northwest, and the moving speed of the cultivated land's center of gravity experienced a process of "slowed (1975-1995)-accelerated (1995-2005)-decelerated (2005-2015)". The spatial distribution pattern of the standard deviation ellipse appeared more obvious in the northwest direction after 1995, and the area of ellipse increased substantially. The growth range of the cultivated land was constantly strengthened and had a trend of expansion to the northern desert area of the basin. Through LISA frequency geo-spectrum analysis, we found that the total proportions of stable invariant and low frequency types accounted for 84.58%, and the spatial pattern of cultivated land was relatively stable. The overall spatial aggregation of the cultivated land generally shifted from a state of "Low-Low" aggregation "Low-High" aggregation to a state of "High-High" aggregation, which showed high-density expansion of cultivated land in the Manas River Basin from 1975 to 2015. By constructing the geo-spectrum of LISA frequency and combining the gravity center model and standard deviation ellipse, spatial evolution trends in the regional cultivated land were comprehensively and dynamically outlined, and the theoretical basis and decision-making foundations for scientific management and effective protection of cultivated land resources were provided.