Soil respiration is an important source of atmospheric CO2, with the potential for large positive feedbacks with global warming. The size of these feedbacks will depend on the relative sensitivity to temperature of very large global pools of highly stable soil organic matter (SOM), with residence times of centuries or longer. Conflicting evidence exists as to the relationships between temperature sensitivity of respiration and stability of SOM, as well as the temperature sensitivity of individual stabilisation mechanisms. This PhD considers the relationship between different stabilisation mechanisms and the temperature sensitivity of SOM decomposition. I used physical fractionation to isolate SOM pools with a variety of turnover rates, from decadal to centennially cycling SOM, in a peaty gley topsoil from Harwood Forest. Mean residence times of SOM as determined by 14C dating was most strongly affected by depth, providing stability on a millienial scale, while OM-mineral associations and physical protection of aggregates provided stability to around 500 years. Chemical characteristics of organic material in these fractions and whole soils (13C CP-MAS NMR spectroscopy, mass spectrometry, FTIR spectroscopy, thermogravimetric analysis, ICP-OES) indicated the relative contribution of different stabilisation mechanisms to the longevity of each of these fractions. Two long-term incubations of isolated physical fractions and soil horizons at different temperatures provided information about the actual resistance to decomposition in each SOM pool, as well as the temperature sensitivity of respiration from different pools. Naturally 13C-labelled labile substrate additions to the mineral and organic horizons compared the resistance to priming by labile and recalcitrant substrates. Manipulation of soil pore water was investigated as a method for isolating the respiration of SOM from physically occluded positions within the soil architecture. Contadictory lines of evidence emerged on the relative stability of different SOM pools from 14C dating, incubation experiments and chemical characterisation of indicators of stability. This led to the interpretation that physical aggregate protection primarily controls SOM stability within topsoils, while mineral and Fe oxide stability provides more lasting stability in the mineral horizon. Less humified and younger SOM was found to have a higher sensitivity to temperature than respiration from well-humified pools, in contrast to predictions from thermodynamics.