The paper presents the application of thermal remote sensing for mapping hydrocarbon polluted sites. This has been achieved by mono-window algorithm for land surface temperature (LST) measurements, using multi-date band 6 data of Landsat Thematic Mapper (TM). The emissivity, transmittance and mean atmospheric temperature were used as critical factors to estimate LST. The changes in the surface emissivity due to oil pollution alter the apparent temperature, which was used as a recognition element to map out oil polluted surfaces. The LST contrast was successfully used to map spatial distribution of hydrocarbon pollution in the Burgan Oil field area of Kuwait. The methodology can be positively used to detect waste dumping, oil spills in oceans and ports, besides environmental management of oil pollution at or near the land surface.