Computational methods based on density functional theory have been applied to a prospective study of dinuclear transition metal complexes that may show strong exchange coupling interactions through very long bridging ligands. The results indicate that M(III) complexes (being M= Cr, Mn or Fe) with dicyanamidobenzene-type ligands are specially promising for this purpose, since strong ferromagnetic or antiferromagnetic coupling is predicted between paramagnetic metal cations at distances as long as 25 A. The existence of ferromagnetic or antiferromagnetic coupling in the complexes with the different isomers of dicyanamidobenzene can be rationalized in terms of molecular orbitals.