The theory of the NMR paramagnetic shift in pseudotetrahedral complexes for Ni(II) and Co(II) has been developed and applied to experimental results for complexes using the same ligands, The theory used the atomic overlap method, AOM, for the ligand field and included the P as well as the F states. It specifically allowed for different spin delocalization from different d orbitals, It has been shown that the shifts in Ni(II) complexes is sensitive to delocalization of spin only from the metal d(+/-1) orbitals. For Co(II) complexes, the shift is mainly due to the dir orbitals but there could be a measurable effect from the d(+/-2) orbitals. It has been further shown that the failure to recognize the differing spin delocalization of different metal d orbitals in previous analyses of paramagnetic shifts has led to estimates of spin density in the ligands to be too small by a factor of 2.5 in Ni(II) complexes and 3.0 in Co(II) complexes. It has also been discovered that the so-called contact shift is not isotropic, particularly for Ni(II) complexes, and this leads to a dipolar contribution for the average contact shift that could be significant for paramagnetic shifts that are primarily due to delocalization of spin in pi aromatic MOs. The theory has been used to examine experimental paramagnetic shifts for Co(II) and Ni(II) complexes