In Midwestern agricultural fields, subsurface tile drainage has been widely used to remove excess water from the soil through perforated tubes installed beneath the ground surface. While it plays an important role in enabling agricultural activities in wet but productive areas, this system is a major driving factor affecting water and nutrient dynamics, and water quality in this region. However, despite its critical role, the specific locations of subsurface tile drainage structures are not generally available nor well captured by conventional optical image processing due to soil surface features, such as topographic depressions and tillage. To overcome these challenges, in this study, we have explored the potential of using thermal images to identify the location of a subsurface drainage pipe. The hypothesis is that the unique spatial distribution of soil moisture set up by tile drains can result in the difference in surface soil temperature between areas near and away from drainage pipes. Toward this objective, we designed and developed an experimental device based on a dimensionless analysis at a scale of 1:20, which was deployed in the open air for 4.5 months. The experimental results demonstrate that (1) there is an ideal time for thermal image acquisition that maximizes the contrast between the regions close to and distant from subsurface drainage systems, and (2) the thermal image processing approach proposed in this study is a promising tool that has advantages of higher accuracy and stability in localizing subsurface drainage pipes over optical image-based approaches.