CO2 capture and storage (CCS) is a technological solution to stabilize or even reduce the atmospheric concentration of the greenhouse gas CO2, to mitigate climate change. In this context, MgO is a promising solid CO2 sorbent, as the energy penalty sorbent regeneration is comparatively small, but it requires the addition of promoters, typically alkali metal nitrates, to yield acceptable kinetics. Under operating conditions, the promoters are in a molten state. The main objectives of this work are (i) to assess experimentally the validity of different reaction mechanisms for the CO2 uptake of promoted MgO that are currently debated in literature and (ii) to elucidate the processes that lead to sorbent deactivation. Our experimental results support the mechanism in which the dissolution of MgO in the molten nitrate promoter is the rate-limiting step for carbonation. We were able to establish a direct correlation between the solubility of MgO in the promoter and the initial rate of carbonation. In addition, a systematic study of a large number of promoter compositions (mixtures of LiNO3, NaNO3, KNO3) indicate that promoters with a lower melting point exhibit higher CO2 uptakes, presumably due to their lower viscosity and, thus, higher ion mobility at a given temperature. Concerning the cyclic stability of promoted MgO, a decay of its CO2 uptake with number of carbonation/calcination cycles is ascribed only partially to sintering. Instead, the surface migration of the promoter was identified as an at least equally relevant deactivation mechanism. Importantly, it was also found that the CO2 uptake of the deactivated sorbent can be restored to a large extent with a simple hydration step.