Ocean iron fertilization (OIF) is being considered as a strategy for mitigating rising atmospheric CO2 concentrations. One model for implementation is the sale of carbon offsets. Model- ing studies predict that OIF has the potential to produce a material difference in the rise of atmos- pheric CO2 over the next several decades, but this could only be attained by alteration of the ecosys- tems and biogeochemical cycles of much of the world's oceans. The efficacy of OIF on this scale has not been proven. However, the consequences of successful implementation must be considered now, for 2 important reasons: (1) to determine if the environmental effects would be predictable and veri- fiable, and if so, acceptable; and (2) to establish whether the basis for valuing carbon offsets — an accurate audit of net reductions in cumulative greenhouse gas potential over 100 yr — can be met. Potential side-effects of widespread OIF that must be considered include a reduced supply of macronutrients to surface waters downstream of fertilized regions, increased emissions of the potent greenhouse gases nitrous oxide and methane, and changes in the extent or frequency of coastal hypoxia. Given the uncertainties inherent in ocean models, predictions of environmental effects must be backed up by measurements. Thus, to go forward with confidence that the effects of rising CO2 could indeed be mitigated through OIF over the next century, and to establish the foundations for auditing carbon offsets, it must be explicitly demonstrated that methods exist to predict and detect downstream effects of OIF against the background of both climate variability and global warming. We propose that until the side-effects of widespread OIF can be shown to be verifiable — and there is good reason to believe that they cannot — OIF should not be considered a viable technology for climate mitigation.