Rising temperatures in urban environments are driven by the Urban Heat Island (UHI) phenomenon, which poses a critical challenge. This systematic review centers on Land Surface Temperature (LST) within urban areas, aiming to uncover predictive models, influential variables, and key methodologies for Land Use and Land Cover (LULC) classification between 2019 and 2023. Employing the PRISMA framework, a comprehensive literature search was conducted. The identification, screening, and keyword-based selection process led to the inclusion of pertinent research. Emphasis was placed on understanding the prediction of LST, including satellite data, factors affecting LST, and the relationships between these factors and prediction methods. The review reveals a surge in research focusing on LST prediction, highlighting the growing importance of understanding urban heat dynamics. Satellite imagery emerged as a pivotal data source, particularly Landsat and MODIS data. Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), and urban indices significantly impacted LST. Furthermore, positive correlations were identified between these factors and machine learning (ML) prediction methods. This systematic review underscores the urgency of comprehending LST in urban contexts. The proliferation of research in LST prediction reflects the increasing recognition of the need for data-driven solutions to mitigate UHI challenges. Identifying cooling factors and their complex interplay with urban heat dynamics is paramount for effective urban planning. Machine learning’s role in predicting LST demonstrates its potential in urban climate studies, providing a basis for evidence-based decision-making and the development of resilient, sustainable urban spaces amidst rising temperatures.