RHF/6-31G* and B3LYP/6-31G* computations were performed on the silene [Me2NN = (CMe)O](2)Si = CH2 (6) and analyzed through the use of properties of atoms in molecules. Three stationary states of 6 belong to its non-chelate (6a) and two intramolecularly N-donor-stabilized forms, five-membered mono-chelate (6b) and previously unknown (for the silenes) bis-chelate (6c), with three-, four-, and five-coordinate doubly bonded silicon, respectively. On going from 6a to 6b and 6c, initially planar silicon attains distorted tetrahedral and square pyramidal structures, whereas the Si = C double bond becomes more polar and changes its distance from 1.674 Angstrom to 1.693 and 1.704 Angstrom and from 1.691 Angstrom to 1.701 and 1.713 Angstrom at the RHF and B3LYP levels, respectively. The RHF and B3LYP N-Si distances in 6b (1.988 and 2.031 Angstrom) are shorter than in 6c (2.187 and 2.140 Angstrom). The N --> Si bond in chelates 6b,c is described as highly polar, but of sufficiently covalent character. The four-center six-electron (4c-6e) model is proposed for the silicon bonding in the N2Si = C moiety of 6c. High energetic advantages of the chelate 6b and 6c forms over 6a (26.8 and 31.4 kcal/mol at the B3LYP/ 6-31G* level including the ZPE correction and 32.4 and 36.3 kcal/mol at the RHF/6-31G* level, respectively) suggest that intramolecular N-donor stabilization may be sufficient to observe silene 6 under relatively mild conditions.