Representation of a system using an average glass transition temperature (T
-g) value will become inadequate if the system is not sufficiently homogene
ous in terms of physical structure and chemical composition. Many food syst
ems are heterogeneous, and therefore a nonuniform distribution of T-g value
s is expected. In this exploratory article, the basis for determination of
T-g using nuclear magnetic resonance (NMR) techniques is discussed, and a m
agnetic resonance imaging (MRI) method is proposed for determining the spat
ial distribution of T-g in food systems. The T-g maps for maltodextrin and
bread samples, obtained using this method, are presented.