Linear polynuclear paddlewheel complexes-"extended metal atom chains" or "metal strings"-have provided attractive models for the study of metal-metal bonding, magnetism and conductivity since their discovery in the 1990s [1]. Their helicoidal chirality, arising from mutual steric hindrance of the 3-pyridyl protons, resulting in the twisting of the equatorial ligand around the metal axis (see figure), has been less studied. Nonetheless, in one of the few examples of chiral resolution, the obtained enantiomers of a trinickel complex showed a remarkably high specific rotation of 5000 deg•mL•g −1 •dm −1 [2], motivating us to seek a general technique for the chiral resolution of such racemates. We have developed a procedure based on anion exchange for the chiral resolution of [M3(dpa)4] 2+ salts (M = Co(II) or Ni(II), Hdpa = 2,2'-dipyridylamine). Homochiral arsenyl tartrate (AsT) salts promoted the selective crystallization of [-M3(dpa)4(MeCN)2](NBu4)2[-AsT]2, or [-M3(dpa)4(MeCN)2](NBu4)2[-AsT]2 in the P4212 space group. The enantiopure compounds demonstrated surprisingly large optical activities using UV-vis, Raman and infrared spectroscopy in solution and, for the cobalt derivatives, in the X-ray range at the Co K-edge in single crystals. An intense X-ray linear dichroism was observed in the orthoaxial crystal orientation, whereas it vanished in the axial confirmation, while the angular dependence of the circular dichroism spectra followed the expected (3cos 2 − 1) function, thus spectroscopically confirming the D4 crystal symmetry. X-ray magnetic circular dichroism and X-ray magnetochiral dichroism signals at the Co K-edge were not detected, likely due to a strongly delocalized spin density on the metal-metal bonded tricobalt core. Nevertheless, these results establish that chiral polynuclear paddlewheel complexes can be cleanly resolved using selective crystallization and demonstrate considerable optical activity in the infrared, UV-vis and X-ray energy ranges, thus potentially offering future perspectives in non-linear optics and asymmetric synthesis [3].