The objective of this study is to map the land cover units (LCU) of the Plateau of Allada from the supervised classification by the maximum likelihood method of Landsat multi-spectral images (TM 1986, ETM+ 2000 and OLI 2020) and to highlight their evolutions by analyzing the areas and the rates of change. The approach used combines image pre-processing, identification of land use unit classes, construction and execution of supervised classification by the maximum likelihood method and quantification of change. All of these treatments applied to the images, allowed us to obtain the land use maps in 1986, 2000 and 2020 with five land use classes (forest/gallery/swamp, plantation, mosaic/fallow/crops, dwellings/bare ground, and water yard). Analysis of the dynamics of the land use units from 1986 to 2020 shows a progressive trend in the area of the mosaic/crop/fallow (25.97%) and habitat/bare soil (10.51%) classes and a regressive trend in the forest/gallery/swamp (-5.82%), plantation (-0.13%) and water (-0.25%) classes. Quantification of change from 1986 to 2020 is assessed by an estimated rate of change (Tc) of -5.82% (forest/gallery/swamp); 1.38% (mosaic/fallow/crop); 2.79% (dwellings/bare ground); -0.13% (plantations) and -0.25% (watercourse). This evolution is due to the expansion of the housing/bare soil and the mosaic/fallow/crops areas. This is the result of the strong human pressure on the vegetation formations. These results constitute a decision-making tool for the sustainable management and urbanisation of the Plateau of Allada.