Radiocarbon differences between benthic and planktonic foraminifera (B-P ages) and radiocarbon projection ages are both used to determine changes of the past ocean circulation rate. A global 3-D ocean circulation model with a constant modern ocean circulation is used to study which method is less influenced by atmospheric Δ¹⁴C variations. Three factors cause uncertainties: first, the long equilibration time of the ocean after atmospheric Δ¹⁴C changes; second, different mixing processes in the ocean, which cause an ocean response of smaller amplitude than the atmospheric forcing; and third, the unknown source region and corresponding initial surface ¹⁴C reservoir age of subsurface waters. The model suggests that BP ages and projection ages have lower uncertainties the closer they are to deepwater formation zones. In the North Atlantic the B-P age method is less influenced by atmospheric Δ¹⁴C variations than the projection-age method. Projections ages vary less in the Pacific as long as atmospheric Δ¹⁴C decreases linearly. A more irregular atmospheric Δ¹⁴C evolution leads to age variations of similar magnitude with both methods. On the basis of the model experiment, we suggest a potential improvement of the projection-age method.