An empirical model was developed to estimate the thermal conductivity of heat-treated straw bulks based on laboratory experiments. During the measurements, two different types of straw were investigated, barley and wheat. Barley was used to composing our empirical model and define the influencing model parameters, and wheat straw was used for validation. Both straws were heat-treated in a dry oxidative ambient in five temperature steps from 60 to 180 °C. The thermal conductivity was measured at 120 kg m−3 bulk density after every treatment cycle. In addition, we were looking for the most suitable measurement methods to detect changes in material structure related to thermal conductivity in the range of relatively low-temperature treatments. Thermogravimetric measurement was conducted, and the mass loss and elemental composition were measured after every treatment cycle. The measurements showed that the mass percentage ratio of carbon in straw increased, and the mass percentage of oxygen decreased in the investigated temperature range. We identified and separated the following parameters of the model, which can estimate the relative thermal conductivity of heat-treated stem bulks: relative residual mass, relative mass percentage ratio of carbon content and oxygen content. We divided the model into two parts, creating a simpler but worse approximation (the measurements required for this are much easier to perform) and a slightly more complex but better approximation. After the validation, our model achieved good agreement with the relative thermal conductivities calculated by the measured thermal conductivities.