Heat transport in the geosphere is important in applications of geothermal energy systems, thermal remediation technologies, and design of energy foundations. The study of these applications would benefit significantly from the ability to collect temperature data from within the porous media system, and at high spatial and temporal resolutions. Temperature measurements made using conventional probes have high temporal resolution but are limited in their spatial resolution. Higher-resolution methods, such as thermal imaging, are limited to measurements of an exposed face of an experiment. This paper presents the development of a technique for measuring temperature using transparent soil. In typical transparent soil, the refractive indices of the soil particles and the pore fluid are matched, creating invisible soil particles when saturated. However, because the refractive indices of the soil particles and pore fluid are different functions of temperature, the degree of transparency decreases as the temperature increases or decreases from the transparency temperature. As such, changes in transparency are detected by digital photographs and can be calibrated and used to measure temperature. This paper presents relationships between temperature and normalized pixel intensity for two oil-fused silica combinations. One combination used an oil mixture with a transparency temperature of 25°C and the other, which was constructed using one of the oils in the mixture, has a transparency temperature of 4°C. The results show that there must be at least a 10°C differential from the transparency temperature to ensure a linear relationship between temperature and normalized pixel intensity. The capabilities of transparent soil constructed with the second oil are displayed in two laboratory experiments, which provide direct comparisons between transparent soil, conventional temperature probes, and thermal imaging. The results show the transparent soil provides reliable temperature fields across the experimental domain at high spatial and temporal resolutions.