Semi-classical transitional state theory and Bell model with a H-tunneling correction are the traditional basis for the understanding of the hydrogen kinetic isotope effects (KIEs). However, they fail to explain many observations in H-transfer reactions. The major reason is that they do not consider the role of the heavy atom motions in the H-transfer chemistry. Marcus-like model is a phenomenological model which treats H as a wave between vibrational donor and acceptor and can explain the observations from temperature independence of KIEs for the wild-type enzymes (∆E_a = E_aD - E_aH = 0) and temperature dependence of KIE’s in mutated enzymes (∆E_a > 0). Within this new model, KIE originates from different H/D wave-function overlap between activated reactant and product at all possible donor-acceptor distances (DAD’s) sampled by the heavy atom motions. In this work, two hypotheses were recognized from this model, which link structure/DAD to KIEs and the temperature dependency of the KIEs. They are, (1) a shorter DAD is required for D-tunneling than for H-tunneling, and (2) the more rigid the reaction centers the weaker will be the temperature dependence of the 1° KIEs, i.e. a smaller ∆E_a. A series of hydride transfer reactions of NADH/NAD⁺ analogs in acetonitrile were designed to systematically investigate the hypotheses. Experiments studied electronic, steric, and temperature effects on kinetics and produced rate constants ranging from 105 to 1.0 M-1s-1, KIE’s from 2.5 to 5.31, and ∆E_a’s from nearly 0 to 1.19 kcal/mol. It was found that H- and D-tunneling processes use different tunneling-ready state electronic structures. It was also found that the ∆E_a increases as the rigidity of the donor-acceptor centers decreases. The latter results replicate the trend of ∆E_a’s observed from enzymes to mutated enzymes and have opened a new research direction that studies structure – ∆E_a relationship. Results support the two hypotheses, and thus are consistent with the Marcus-like H-tunneling model.