Nitrogen activation by nitrogenase is one of the most important enzymatic processes on earth. In spite of the determination of X-ray structures of increasingly higher resolution, the nitrogenase mechanism is still not understood. In the most recent X-ray structures it has been shown that a carbon resides in the center of the MoFe-cofactor. Its role is not known. Recent spectroscopic studies, mainly EPR, have come closest to obtaining a molecular mechanism for activating nitrogen. Two hydrides have been shown to play a key role in this context. In the present study, the mechanism for nitrogenase has been investigated by hybrid DFT using a cluster model. This approach has been shown to be very successful for predicting mechanisms for other redox-active enzymes, such as the one for photosystem II, but has so far not been used in its most recent form for nitrogenase. The mechanism obtained has large similarities to the one suggested by spectroscopy, with a reductive elimination of two hydrides just before nitrogen binding. However, a very surprising finding is that the central carbon becomes protonated and has to move out of the cavity as a methyl group before the hydrides can be formed. This has not been suggested before.