Knowledge of the uptake of atmospheric carbon dioxide by the North Atlantic is important in understanding the global carbon budget. Obtaining an accurate value for the ocean sinks in the northern hemisphere temperate zone, in particular, would make it possible to be more quantitative about the enigmatic land sink in these latitudes. The global ocean sink is dominated by the North Atlantic, despite its small area in comparison with the North Pacific. Of the possible methods for calculating the uptake, potentially the most informative is that based on the gas exchange equation because information about the seasonal and spatial trends can be obtained. However, there still remain serious questions about which form of the gas exchange coefficient is appropriate to carbon dioxide. Here we summarize current knowledge of the gas exchange coefficient, including recent new evidence which supports lower gas exchange rates and which, therefore, generally supports the Liss-Merlivat prediction for less soluble gases. In view of the uncertainties still surrounding the direct calculation, we investigate methods for calculating the basin-wide uptake by exploiting what is known about the general circulation of the North Atlantic. The uptake can be estimated by dividing it into pre-industrial steady state and anthropogenic contributions. We make a new estimate of the pre-industrial flux based on the heat budget of the North Atlantic, and also consider two earlier calculations of the same quantity. Taking a best value, adding the better-known anthropogenic flux and making small corrections, we find a value of 0.7± 0.15 Gt C a
-1
for the uptake of the North Atlantic, north of 15° N, in the mid 1980s. Though larger than the gas exchange based estimates of Tans
et al.
(1990), this value is not enough to obviate the need which they deduced for a sizeable ‘land-based’ northern hemisphere sink for CO
2
.
Isopycnic modeling of the North Atlantic heat budget