A linear dilatometer was used to investigate the effects of cure conditions, mold types, and the presence of filler in a model epoxy. These studies revealed shrinkage in the cured epoxy when heating it through the glass transition region. The magnitude of the shrinkage, related to stress buildup in the epoxy during cure, was influenced by the epoxy processing conditions, filler content, and the nature of the mold used to contain the resin. Curing at a lower temperature, prior to a post-cure, decreased the magnitude of observed shrinkage. Annealing samples outside the mold allowed stresses that developed during the cure process to be reduced and led to less shrinkage observed than for samples annealed in the mold. Samples cured in a silicone mold exhibit less shrinkage in the dilatometer than samples cured in an aluminum mold. Soaking samples in water prior to dilatometric analysis led to stress relaxation in the sample and decreased the observed shrinkage. Specimens containing titanium dioxide filler exhibited less shrinkage than unfilled samples. The filled samples also exhibited less expansion slightly above the glass transition temperature. The reduced expansion might arise from restricted chain mobility in the presence of the filler.