The sterically hindered, three-coordinate metal systems M[N(R) Ar](3) (R = Bu-t, Pr-i; Ar = 3,5-C6H3Me2) are known to bind and activate a number of fundamental diatomic molecules via a [Ar(R)N](3)M-L-L-M[N(R)Ar](3) dimer intermediate. To predict which metals are most suitable for activating and cleaving small molecules such as N-2, NO, CO, and CN-, the M-L bond energies in the L-M(NH2)(3) (L = O, N, C) model complexes were calculated for a wide range of metals, oxidation states, and d(n) (n = 2-6) configurations. The strongest M-O, M-N, and M-C bonds occurred for the d(2), d(3), and d(4) metals, respectively, and for these dn configurations, the M-C and M-O bonds were calculated to be stronger than the M-N bonds. For isoelectronic metals, the bond strengths were found to increase both down a group and to the left of a period. Both the calculated N - N bond lengths and activation barriers for N-2 bond cleavage in the (H2N)(3)M-N-N-M(NH2)(3) intermediate dimers were shown to follow the trends in the M-N bond energies. The three-coordinate complexes of Ta-II, W-III, and Nb-II are predicted to deliver more favorable N-2 cleavage reactions than the experimentally known Mo-III system and the (ReTaIII)-Ta-III dimer, [Ar(R)N](3)-Re-CO-Ta[N(R)Ar](3), is thermodynamically best suited for cleaving CO.