THE iron hypothesis(1-3)-the suggestion that iron is a limiting nutrient for plankton productivity and consequent CO2 drawdown-has been tested by small-scale experiments in incubation bottles in the subarctic Pacific(2,4) and Southern(5-7) Oceans, and by a recent large-scale experiment in the equatorial Pacific Ocean(8,9). Here we test the idea by looking at natural levels of productivity in regions of the Southern Ocean with differing iron abundance, In the southerly branch of the Antarctic circumpolar current (ACC), upwelling of deep waters supplies sufficient iron to the surface to sustain moderate primary production but not to permit blooms to develop, In contrast, within the fast-flowing, iron-rich jet of the polar front (PF), spring blooms produced phytoplankton biomass an order of magnitude greater than that in southern ACC waters, leading to CO2 undersaturation. The plankton-rich PF waters were sharply delineated from adjacent iron-poor waters, indicating that iron availability was the critical factor in allowing blooms to occur.