Land use and land cover (LULC) change and the associated risk of soil erosion have become a global environmental problem that threatens the sustainability of land resources. This work analysed the LULC dynamics and estimated the soil erosion risk potential of Nkonjwa community in Eswatini using GIS, remote sensing and Revised Universal Soil Loss Equation (RUSLE) model. Landsat imagery for the years 1990, 1999, 2009 and 2019 were classified using maximum likelihood classification, and the produced LULC maps used to analyse LULC changes in the area over the 29 years. The area was classified into water, agricultural land, forests and shrub and bare land major LULC classes. Resultant analysis indicated a significant shift from forests, water and agricultural land to bare land, with a decline of 16.4%, 4.2% and 34.3%, respectively. Increased bare land implied an increased vulnerability of the area to land degradation by soil erosion. A soil erosion map generated using the RUSLE model showed that severe, very high, high, moderate, low and very low erosion classes accounted for 1.3%, 0.9%, 2.6%, 22.2%, 11.8% and 61% of the total area, respectively. Average annual soil loss from each land cover class ranged from 2.9 to 6.9-ton ha-1 yr-1, with the highest rates on bare land. The study concluded that over the years, there has been change in the LULC of Nkonjwa area and the LULC change analysis and spatial RUSLE model parameters can serve as effective inputs in deriving strategies for conservation planning in communities of Eswatini.