A recent experimental dtermination[1] of the dissociation energies (Do) for H2N-H, H2N+-H and H2N-H+, the ionization energies for NH3 and NH2 resulted inlarge deviations when compared with those of the earlier values and the QCISD(T)/6-311+G(3df,2p) ab initio calculations. We have performed some higher level ab initio calculations on these data by utilizing the Gaussian 92/DFT and Gaussian 94 pakages of programs and have assessed the available experhaental values. Our calculations were carried out at the QCISD (TQ)/aug-cc-pVDZ, G2(QCI), QCISD(T)/6-311 ++G(3df,3p(1)and QCISD(T)/aug-cc-pVTZ levels of theory. Geometries were optimized at both of the MP2(full)/6-31G(d) and the MP2(full)/6-31(d,p) levels, and were compared with those of the expcriments if available. The MP2(full)/6-31G(d,p) tight-optimized geometries for the neutrals are closer to those of the experiments than those of th.MP2 (full)/6-31G(d), and are in excellent agreoment with the experimental results as shown in Table 1. In this case,we assumed that the optimized geometries for the cations would be better if p polarization functions are added to the hydrogen atoms. We firstly noted that the sym-mtry of the NH3+ cation was D3h, other than Cs. as reported in ref.[1]. All of the zero-point energies and the final geometries are calculated at the MP2(full)/6-31G(d,p) level of theory- We have also repeated the QCIS D(T )/6-311 + G(3df,2p) calculations of ref. [1], because we could not identify their level of goemetry optithezation.It is found that the total energy, -55.244 19 Hartrees, for NH (1A1 ) in ref.[1] might be in error. Our result is -55.336 29 Hartrees at the same level of theory. At our highest level [QCISD(T)/aug-cc-pVTZ] of calculations as shown in Table 3, the D0, (temperature at zero Kelvin) values of H2N-H, H:N+-H(3B1for NH2+ ) and H2N- H+ are 4.51, 5.49 and 8.00 eV. respectively. These data reported in re f.[1] were 4.97, 5.59 and 8.41 eV, respectively. Our result on Do(H2N-H) supports the work of ref.[2,3,5,6]. The ionization energies (I