The iron-molybdenum cofactor (FeMoco) is responsible for dinitrogen reduction in Mo nitrogenase. Unlike the resting state, E 0 , reduced states of FeMoco are much less well characterized. The E 2 state has been proposed to contain a hydride but direct spectroscopic evidence is still lacking. The E 2 state can, however, relax back the E 0 state via a H 2 sidereaction, implying a hydride intermediate prior to H 2 formation. This E 2 !E 0 pathway is one of the primary mechanisms for H 2 formation under low-electron flux conditions. In this study we present an exploration of the energy surface of the E 2 state. Utilizing both cluster-continuum and QM/MM calculations, we explore various classes of E 2 models: including terminal hydrides, bridging hydrides with a closed or open sulfide-bridge, as well as models without. Impor-tantly, we find the hemilability of a protonated belt-sulfide to strongly influence the stability of hydrides. Surprisingly, nonhydride models are found to be almost equally favorable as hydride models. While the cluster-continuum calculations suggest multiple possibilities, QM/MM suggests only two models as contenders for the E 2 state. These models feature either i) a bridging hydride between Fe 2 and Fe 6 and an open sulfide-bridge with terminal SH on Fe 6 (E 2 -hyd) or ii) a double belt-sulfide protonated, reduced cofactor without a hydride (E 2 -nonhyd). We suggest both models as contenders for the E 2 redox state and further calculate a mechanism for H 2 evolution. The changes in electronic structure of FeMoco during the proposed redox-state cycle, E 0 !E 1 !E 2 !E 0 , are discussed.