The importance of sea ice in polar carbon budgets is currently not known. The concept of a sea-ice carbonate pump, where the expulsion of CO₂-rich brine during sea-ice growth in the fall and winter, and melt of (CO₂) undersaturated ice during the spring and summer, has recently been put forward as a potentially important process contributing toward the uptake of atmospheric CO₂ by the surface water. The recent finding of CaCO₃ in natural sea ice also raises the possibility that the dissolution of the mineral during sea-ice melt could augment CO₂ uptake of the surface waters independently of biological activity during the spring and summer season. However, there remain several important unknowns in our understanding of the ocean–ice–atmosphere cycling of CO₂, including the seasonal patterns and controls over the CO₂ flux between sea ice and the polar atmosphere. There is increasing evidence from studies in both northern and southern polar seas that sea ice is an active participant in air–ocean CO₂ exchange. Here we present results from two sea-ice experiments over seasonal sea ice in the Arctic, where the CO₂ flux was monitored in conjunction with measurements of the surface’s heat budget of landfast sea ice, in addition to the snow and sea-ice temperature structure and relevant physical properties. We observed air–sea-ice CO₂ exchange throughout the winter, spring and summer seasons and found recurring patterns in the flux (both efflux and uptake) associated with both near-surface heating and ventilation of the snow on both seasonal and diurnal timescales. Fluxes generally ranged between±±1 μmol m^–2 s^–1 (negative flux denotes uptake), although at times uptake and emission were far (by a factor of 3) in excess of this range. The strongest period of uptake (< –3 µmol m^–2 s^–1) is observed in the late spring and corresponds to rapid brine-draining events. The strong temperature dependence demonstrated in our flux dataset leads us to believe that the flux is responding to the phase equilibrium of CaCO₃ in brine at the snow base and near to the ice surface. The subsequent strong uptake maybe the response to sea-water CO₂ undersaturation associated with a combination of photosynthetic uptake, input of meltwater low in dissolved CO₂, and/or the dissolution of CaCO₃. Although these results do not directly corroborate the presence of the recently proposed sea-ice carbonate pump, they do demonstrate a strong influence on air–surface exchange by processes originating within the snow and sea-ice layer, thereby adding to the body of evidence that sea ice is an active participant in air–ocean CO₂ exchange within sea-ice-dominated polar waters.