<p>Land use and land cover (LULC) dynamics are crucial indicators of human activities altering the earth’s surface, leading to environmental changes and land degradation. Ethiopia's rapid population growth, agricultural expansion, deforestation, and urbanization have significantly impacted LULC over the past decades. This study examines the spatiotemporal patterns of LULC dynamics and their relationship with Normalized Difference Vegetation Index (NDVI) trends in the Middle Omo-Gibe River (MOGR) basin. The study employs a machine learning (ML) technique on Google Earth Engine (GEE) to analyze Landsat images from 1990 to 2020, incorporating field observations and participatory methods to capture local perceptions and validate LULC classification and trends. From 1990 to 2020, the MOGR basin experienced significant growth in settlements, agricultural land, and bare land, with varying patterns across decades. Meanwhile, bare land expanded the most (9% annually), followed by agricultural land and settlements from 1990 to 2000. Likewise, settlements proliferated, while shrublands and grazing land decreased due to the expansion of agriculture and urbanization from 2000 to 2010. The construction of Omo-Gibe Reservoir III significantly increased water bodies (from 25.61 to 194 km²) while declining forest land by 22.1% and grazing land by 7.57% from 2010 to 2020. Moreover, the NDVI is increasing, particularly in the central and eastern regions, driven by reforestation initiatives like Ethiopia’s “Green Legacy initiative.” These findings, corroborated by field observations and local perceptions, underscore the urgency of adopting appropriate land use practices and environmental management strategies to ensure sustainable resource utilization and development in the river basins of Ethiopia.</p> Graphical abstract <p></p>