It has been postulated that the heat stabilization of the essential macromolecules in the core of the spore may be produced by dehydration at two levels: (i) the spore is drier at high relative humidity than the vegetative cell and (ii) the core of the spore may be less hydrated than the cortex and the coat. The latter postulate was subjected to experimental testing by 1H-NMR studies of the water signal in the five species of spores and coat and (coat + cortex) preparations. The transverse relaxation rate ( 1 T 2 ) was determined in samples equilibrated at constant relative humidity. To allow for the effect of paramagnetic ions on 1 T 2 a model system (wool keratin) was studied in the presence of known amounts of Ca(II), Mn(II), Cu(II), Ni(II) and Fe(III). Because of the dominant effect of Mn(II) on 1 T 2 , the effect of small amounts of other metal ions in spores was neglected. The relaxation rate of water at a particular relative humidity and manganese concentration was consistently less for intact spores than for coat or coat + cortex, hence the water in the core is more mobile than in the outer integuments. Sorption isotherm studies have shown that at a particular relative humidity there is about as much water in the core as in the cortex and coat. These two results taken together indicate that the hypothesis that the core is drier than the cortex and coat is incorrect, hence the spore is not heat-stabilized in this way. A theory is proposed in which heat stabilization is attributed to immobilization of essential enzymes and nuclei acids by a solid support, calcium dipicolinate, in a similar fashion to the heat stabilization of enzymes in a charged polymer matrix. It is proposed that stabilization is effected by electrostatic and hydrogen bonds between the calcium dipicolinate and the essential macromolecules. Experiments in progress show that enzymes and DNA are heat-stabilized in vitro by calcium dipicolinate.