Two-dimensional thermocline ventilation models for the temperate North Atlantic with differing circulation patterns were calibrated to yield a tritium distribution similar to that observed during the GEOSECS survey. These models were then run for 3He, bomb-produced 14C, radiokrypton, and freons. They were also run for the uptake of fossil fuel CO2. While the models differ significantly in their ability to match the observed 3He and 14C distributions, these differences are not large enough to clearly single out one model as superior. This insensitivity of tracer-to-tracer ratio tomodel design is reflected by the near identity of the fossil fuel CO2 uptake by the various models. This result suggests that the uptake of CO2 by the sea is limited more by the rates of physical mixing within the sea than by gas exchange across the sea surface. If so, then the hope that models employing outcropping isopycnals will enhance the CO2 uptake by the sea and thereby lead to a narrowing in the gap that exists for anthropogenic CO2 budgets is not well founded. The interim strategy of using reservoir models calibrated by tracer distributions appears to be sound.