<p>If dark matter carries a baryon number of two, neutron-antineutron oscillations could require its presence to manifest themselves. If it is in addition very light, in the micro-eV range or up to a few orders of magnitude below, these oscillations could even exhibit a Rabi resonance. Though the magnetic tuning required to convert a macroscopic number of neutrons into antineutrons is not realistic, sizeable enhancements remain possible. Building on this observation, axionic realizations for this scenario are systematically analyzed. For true QCD axion models, we find that the Goldstone boson nature of the axion imposes the presence of axionless <InlineEquation ID="IEq1"> <EquationSource Format="MATHML"><math display="inline"> <mi>n</mi> <mo>−</mo> <mover accent="true"> <mi>n</mi> <mo stretchy="true">¯</mo> </mover> </math></EquationSource> <EquationSource Format="TEX">\( n-\overline{n} \)</EquationSource> </InlineEquation> mixing effects, either in vacuum or in decays, which are sufficiently constrained experimentally to leave no room for axion-induced oscillations. Thus, a generic scalar or axion-like dark matter background would have to exist to induce resonant <InlineEquation ID="IEq2"> <EquationSource Format="MATHML"><math display="inline"> <mi>n</mi> <mo>−</mo> <mover accent="true"> <mi>n</mi> <mo stretchy="true">¯</mo> </mover> </math></EquationSource> <EquationSource Format="TEX">\( n-\overline{n} \)</EquationSource> </InlineEquation> oscillations. Yet, if Nature has taken that path to relate dark matter and baryon number violation, the experimental signature would be striking and certainly worth pursuing.</p>