Twelve experiments with two coarse resolutions of a global ocean model using a variety of surface forcings are analyzed to address the question of what sets the mean transport through Drake Passage. Seven of the experiments do not have an active sea-ice model, but the remaining five do. Previous theories have suggested that the Drake Passage transport is governed by the Cape Horn Sverdrup transport or, alternatively, is proportional to the square root of the meridional Ekman transport at the latitude of Drake Passage. The results presented here do not support either of these theories. The Drake Passage transport depends quite strongly on the isopycnal diffusivity parameter in the model and less strongly on the background vertical diffusivity and horizontal viscosity parameters. However, when the magnitudes of these parameters are fixed, the results show a very strong correlation between Drake Passage transport and both the strength of the meridional Ekman transport at the latitude of Drake Passage and the thermohaline circulation off the Antarctic shelf. The relationships are monotonic, but not linear. The best estimate is that the meridional Ekman transport drives ∼100 Sv of Drake Passage transport, while the remaining 30 Sv are driven by the global thermohaline circulation.