A series of epoxy networks were synthesized in which the molecular weight between crosslinks (M-c) and crosslink functionality were controlled independent of the network chain backbone composition. The glass transition temperature (T-g) of these networks was found to increase as M-c decreased. However, the rate at which T-g increased depended on crosslink functionality. The dependency of M-c on T-g is well described by two models, one based on the concept of network free Volume while the other model is based on the principle of corresponding states. Initially, neither model could quantitatively predict the effect of crosslink functionality in our networks. However, our tests indicated that both the glass transition and the rubbery moduli of our networks were dependent on M-c and crosslink functionality, while the glassy state moduli were independent of these structural variables. The effect of crosslink functionality on the rubbery modulus of a network has been addressed by the front factor in rubber elasticity theory. Incorporation of this factor into the glass transition temperature models allowed for a quantitative prediction of T-g as a function of M-c and crosslink functionality.