This paper presents the results of exact calculations of neutron-deuteron scattering with a non-central, spin-dependent, separable two-body force. The two-body force has just two separable terms, corresponding to the bound deuteron and the antibound singlet deuteron. The model is therefore that obtained by the pole-dominance argument of Lovelace, or the equivalent one-nucleon-exchange approach of Amado, and the only non-central part of the two-body force is that needed to produce a bound deuteron with the correct ( 3S 1+ 3D 1) structure. The calculations, over a range of 3–23 MeV for the neutron laboratory energy, yield vector polarizations that are very small compared with experiment, and of the wrong shape. On the other hand, the second-rank tensor polarization moments of the deuteron are in qualitative agreement with experiment. Thus the model is not adequate for explaining vector polarizations, but apparently is at least qualitatively adequate for second-rank polarization moments, as well as (as is already known) for differential cross sections. This is plausibly connected with the fact, discussed herein, that the simple one-nucleon-exchange process (without iterations) produces secondrank tensor polarizations (and differential cross sections) of substantial magnitude, whereas it produces no vector polarization at all.