This paper used RS and GIS techniques to detect and monitor LULC changes and urban sprawl in Islamabad, Pakistan for the period of 1990, 2000, 2010 and 2020. It analyzed multi-spectral Landsat data by utilizing the Maximum likelihood classification (MLC) algorithm to classify the satellite data of Landsat 4-5 ™, Landsat-7 ETM+ and Landsat-8 OLI/TIRS in raster form collected from USGS for the period between 1990 and 2020. The objective of the study was to classify LULC types, detect changes in each class and measure the rate of urban sprawl. The study employed Arcmap software to obtain all these objectives. The accuracy assessment of all the classified maps was assessed by constructing an error matrix and calculating overall accuracy and kappa coefficient. It then calculated change detection by using overlay procedure and urban sprawl by applying built-up index (BU) technique and by clipping built-up areas to extract and create a map of urban sprawl. Both pre-classification and post-classification techniques were used to classify and detect changes. The result showed that the classified map of 1990, 2000, 2010 and 2020 had overall accuracy of 91%, 91%, 86% and 94%. Similarly these maps had kappa coefficient values of 90%, 89%, 83% and 93%. The classification result of 1990 to 2020 depicted that built-up area increased by 18% while the decline trend was found in forest land 4.92%, Agricultural land 10.31%, barren land 2.17% and water bodies 0.03%. Furthermore, the urban sprawl map showed compact sprawl in the center and scattered sprawl in the periphery areas as urban sprawl increased from 56.93 sq. km in 1990 to 214.46 sq. km in 2020, and had a substantial increase of 157.93 sq. km from 1990 to 2020. The policy lessons drawn from this research suggested that vertical development should be encouraged in Islamabad and CDA bylaws should be flexible to promote vertical development and to restrict sprawl. Future research should be conducted in better urban management using RS and GIS. Supervisor:- Aqeel Anwar