Ocean iron fertilization is being considered as a strategy for mitigating the buildup of anthropogenic CO2 in the atmosphere. Assessment of this strategy requires consideration of its unintended consequences, such as an enhancement of ocean N2O emissions. This feedback could offset the radiative benefit from the atmospheric CO2 reduction significantly, because N2O is a much more powerful greenhouse gas than CO2 itself. Our model results show that the magnitude of this offsetting effect is substantial, but is highly dependent on the location and duration of fertilization. We find the largest offsets (of the order of 100%) when fertilization is undertaken in the tropics, particularly when it is of limited duration and size. Smaller, but still substantial effects are found when fertilization is undertaken elsewhere and over longer periods. These results suggest that any assessment of ocean fertilization as a mitigating option is incomplete without consideration of the N2O feedback.