Termites are important soil dwelling insects known for their severe economic losses to agriculture, forestry, timber and construction sectors. Accurate estimation of termite colony population is critical for evaluating the efficacy of management strategies at a given site. Present study involved estimating the population of a fungus growing termite, Odontotermes spp. at three contrasting sites (urban, semi-urban, and agricultural areas) in District Multan, Pakistan. For this purpose, two contrasting techniques were applied: mark-recapture and constant removal methods (regression and maximum likelihood). The results indicated that mark-recapture yielded higher estimates at agriculture site (1851376) followed by urban (948153.45) and semi-urban site (737816.73). The constant removal technique yielded higher population estimates at semi-urban site (16932.36 by regression method and 10118.73 by maximum likelihood method) while the lowest estimates (9580.67-13670) were calculated at agriculture site by both constant removal methods. The findings will be helpful to explore population distribution pattern across different ecological zones and seasons, and study implications for termite management at different sites.