Three new NiII/ZnII heterometallics, [NiZnL'2(OMe)Cl]2 (1), [NiZnL''(Dea)Cl]2·2DMF (2) and [Ni2(H3L''')2(o-Van)(MeOH)2]Cl·[ZnCl2(H4L''')(MeOH)]·2MeOH (3), containing three-dentate Schiff bases as well as methanol or diethanolamine (H2Dea) or o-vanillin (o-VanH), all deprotonated, as bridging ligands were synthesized and structurally characterized. The Schiff base ligands were produced in situ from o-VanH and CH3NH2 (HL'), or NH2OH (HL"), or 2-amino-2-hydroxymethyl-propane-1,3-diol (H4L'''); a zerovalent metal (Ni and Zn in 1, Zn only in 2 and 3) was employed as a source of metal ions. The first two complexes are dimers with a Ni2Zn2O6 central core, while the third compound is a novel heterometallic cocrystal salt solvate built of a neutral zwitterionic ZnII Schiff base complex and of ionic salt containing dinuclear NiII complex cations. The crystal structures contain either centrosymmetric (1 and 2) or non-symmetric di-nickel fragment (3) with NiNi distances in the range 3.146-3.33 Å. The exchange coupling is antiferromagnetic for 1, J = 7.7 cm-1, and ferromagnetic for 2, J = -6.5 cm-1 (using the exchange Hamiltonian in a form Ĥ = Jŝ1ŝ2). The exchange interactions in 1 and 2 are comparable to the zero-field splitting (ZFS). High-field EPR revealed moderate magnetic anisotropy of opposite signs: D = 2.27 cm-1, E = 0.243 cm-1 (1) and D = -4.491 cm-1, E = -0.684 cm-1 (2). Compound 3 stands alone with very weak ferromagnetism (J = -0.6 cm-1) and much stronger magnetic anisotropy with D = -11.398 cm-1 and E = -1.151 cm-1. Attempts to calculate theoretically the exchange coupling (using the DFT "broken symmetry" method) and ZFS parameters (with the ab initio CASSCF method) were successful in predicting the trends of J and D among the three complexes, while the quantitative results were less good for 1 and 3.