Only recently, the radioactive gas thoron ((220) Rn) and its decay products have been regarded as significant health risk in the indoor environment. This is because of new findings of increased thoron concentrations in traditional mud dwellings and considerations leading toward reduced action levels for natural airborne radionuclides. A model which describes the sources and sinks of thoron and its decay products should help to assess the indoor exposure. This work presents an extensive depiction of the influences of indoor conditions on the occurrence of these radionuclides. Measurements were performed in an experiment room and in mud dwellings in China and India. Mud even with an average (232) Th concentration was identified as a significant thoron source. The spatial distribution of the decay products proved to be homogeneous, which is in contrast to thoron gas. The prominent contribution of the unattached and attached decay product (212) Pb to the exposure was elaborated. The theoretically derived impact of air exchange and aerosol concentration, which determines the proportion of unattached decay products, could be confirmed. Transfer coefficients of the model were determined. The thoron model with these transfer coefficients predicts annual doses of almost 2 mSv for dwellers of traditional Chinese and Indian mud buildings, confirming the potential health impact of thoron. PRACTICAL IMPLICATIONS: The radioactive noble gas radon with its decay products is well known as a health risk. After increased concentrations of the isotope (220) Rn (thoron) have been found in traditional Chinese mud-walled cave dwellings, the need for a model that describes the occurrence of thoron and its decay products indoors has arisen. This work presents such a model from the emergence of thoron in the building material until the decay to the stable (208) Pb and discusses the various influences on the occurrence of these nuclides. The model makes possible to predict the exposure of people staying in a room to thoron and its decay products and - combined with a dose model - to calculate their inhalation doses from easily measurable data.