Non-aqueous magnesium-oxygen (or ‘Mg-air’) batteries are attractive next generation energy storage devices due to their high theoretical energy densities, low cost, and potential for rechargeability. Prior experiments identified magnesium oxide, MgO, and magnesium peroxide, MgO₂, as the primary discharge products in a Mg/O₂ cell. Charge transport within these nominally-insulating compounds is expected to limit battery performance, yet these transport mechanisms are either incompletely understood (MgO₂) or remain a matter of debate (MgO). The present study characterizes the conductivity associated with intrinsic (point) defects within both compounds using first-principles calculations. For MgO, negative Mg vacancies and hole polarons localized on oxygen anions were identified as the dominant charge carriers. Nevertheless, the large formation energies associated with these carriers suggest that their equilibrium concentrations are low. Furthermore, a large asymmetry exists in the carrier mobility: Mg vacancies are essentially immobile at room temperature, while hole polarons are highly mobile. Importantly, the calculated formation and mobility data for MgO are in remarkable agreement with the three “Arrhenius branches” observed in conductivity experiments, and thus clarify the long-debated transport mechanisms within these regimes. In the case of MgO₂, electronic charge carriers alone – electron and hole polarons – are the most prevalent. Similar to MgO, the concentration of carriers in MgO₂ is low, and poor mobility further limits conductivity. We conclude that: (i.) sluggish charge transport in MgO or MgO₂ will limit battery performance when these compounds cover the cathode support, and (ii.) what little conductivity exists is primarily electronic in nature (i.e., polaron hopping). Artificially increasing the carrier concentration is suggested as a strategy for improving battery performance.