Determination of soil constituents and structure has a vital role in agriculture generally. Methods for the determination of soil carbon have in particular gained greater currency in recent times because of the potential that soils offer in providing offsets for greenhouse gas (CO2-equivalent) emissions. Ideally, soil carbon which can also be quite diverse in its makeup and origin, should be measureable by readily accessible, affordable and reliable means. Loss-on-ignition is still a widely used method being suitably simple and available but may have limitations for soil C monitoring. How can these limitations be better defined and understood where such a method is required to detect relatively small changes during soil-C building? Thermogravimetric (TGA) instrumentation to measure carbonaceous components has become more interesting because of its potential to separate carbon and other components using very precise and variable heating programs. TGA related studies were undertaken to assist our understanding in the quantification of soil carbon when using methods such as loss-on-ignition. Combining instrumentation so that mass changes can be monitored by mass spectrometer ion currents has elucidated otherwise hidden features of thermal methods enabling the interpretation and evaluation of mass-loss patterns. Soil thermogravimetric work has indicated that loss-on-ignition methods are best constrained to temperatures from 200 to 430 °C for reliable determination for soil organic carbon especially where clay content is higher. In the absence of C-specific detection where mass only changes are relied upon, exceeding this temperature incurs increasing contributions from inorganic sources adding to mass losses with diminishing contributions related to organic matter. The smaller amounts of probably more recalcitrant organic matter released at the higher temperatures may represent mineral associated material and/or simply more refractory forms.