The North Atlantic is characterized by strong seasonality in mixed layer depths, resulting in winter recharge of surface layer nutrients and the spring phytoplankton bloom. This is the classical textbook model of seasonal cycles of oceanic biogeochemical processes, but in fact the North Atlantic is the exception rather than the rule. In much of the temperate and subpolar regions of the basin, the vernal accumulation of biomass is accompanied by a marked drawdown of inorganic carbon in the water column and pulses of particle flux to the seafloor. In the classical model, the decline of the C0 2 is balanced by accumulation of biogenic carbon and particle export. The main export mechanisms include sinking of ungrazed but possibly senescent phytoplankton and zooplankton grazing and egestion. Carbon budgets based on observations from the Joint Global Ocean Flux Study North Atlantic Bloom Experiment and Bermuda Atlantic Time Series cannot be closed using the elements of the classical model. That is, the C0 2 drawdown cannot be balanced by biomass accumulation and exports estimated by sediment traps. There are at least three possible routes toward reconciliation: (i) trap estimates are in error and systematically biased; (ii) spatial variability aliasses the observations making budgeting impossible without recourse to coupled three-dimensional models; and/or (iii) the classical model must be abandoned and replaced by a concept in which the accumulation and export of dissolved organic carbon assumes a major role in the North Atlantic carbon balance. At Bermuda, where the most complete data set exists, the weight of the evidence favours the first and third possibilities.