The marine organic carbon cycle was calculated for three different regions in the Atlantic Ocean. Supported by a Geographic Information System (GIS) a new method of spatial modelling was established to enhance detailed resolution, but still keeping the possibility for a basin wide calculation. For a realistic estimation of the world wide marine carbon cycle it is important to account regional aspects because of the high variability of oceanic regions. Therefore, summing up more realistic regional budgets may cause a considerable revision of existing global estimations.In this study, the investigated regions are characteristic for three different marine regimes, which are: the northern North Atlantic with its subpolar climate and deep water formation, the northern West Atlantic with the Gulf Stream and the oligotrophic Sargasso Sea, and the equatorial East Atlantic with its equatorial and coastal upwelling.The flux of organic carbon from the photic zone to the deep sea and its degradation at the seafloor was modelled and extrapolated by detailed empirical analyses of representative data sets for each region. For a spatial extrapolation three representative master parameters were defined: Primary Production (PP in g C m-2 a-1) at the sea surface, organic carbon flux (FCorg in g C m-2 a-1) at the seafloor and the corresponding water depth (z in meters). Organic carbon fluxes at the seafloor were determined by oxygen demand measurements in surface sediments. For each of the regions, available data were collected and a final representative methodically homogenous data set of oxygen fluxes was installed in order to estimate the amount of organic carbon flux at the sediment-water interface. In contrast to these locally limited one-dimensional data, PP at the sea surface derived from satellite imagery and water depth were chosen as two-dimensional highly resolved spatial data sets.Due to the continuous degradation of organic matter in the water column empirical relationships were established for each region by a non-linear regression method. The flux of organic carbon at the seafloor that passes the sediment-water interface was expressed by the product of PP as the initial gross amount of organic carbon at the sea surface and the corresponding water depth:FCorg = PPb * z-c. The quality of this approach was proven by correlation coefficients between 0.89 and 0.96 as well as convincing results obtained by tests of the final regressions with independent data sets. With the help of these regional relationships the spatial distribution of organic carbon flux that reaches the sediment-water interface could be extrapolated for each region. Subsequently, the two gridded spatial data sets of PP by Antoine et al. (1996) and water depth by ETOPO5 (1988), both reflecting the natural distribution, enabled a cell-by-cell-calculation with an adequate resolution of 9.3 * 9.3 km. Equal area projections and overlay-techniques enabled exact planimetric analyses and mass balance calculations of each data set.In order to investigate the benthic-pelagic coupling of the marine organic carbon cycle in each region the gridded flux data sets were divided by GIS-techniques into different zones: biogeographical provinces, morphological structures such as basins or slopes, and depth-dependent zones. Specifically, the northern North Atlantic reveals benthic remineralisation fluxes half as high as the two other regions and also smaller Export Ratios with a spatial variation of 1.7 to 2.0 % at 1000m water depth. The biogeographic provinces of the Gulf Stream and the Sargasso Sea in the northern West Atlantic are divided by different Export Ratios of 3 and 6 % at 1000 m water depth. Compared with global calculations, fluxes twice as high at the continental upper northeast American slope could be distinguished from extremely low fluxes in the deep sea. In contrast to the other regions, the equatorial East Atlantic shows significantly higher benthic fluxes with a constant high level throughout down to the abyssal plain due to the equatorial upwelling and high PP. Thus, local variability and specific dynamics are captured and adequately calculated both depth dependent and on a small regional scale. This partitioning shows a significant reflection of the biogeographical provinces of the sea surface from the benthic point of view. In contrast to global approaches, all three regions show different depth-related benthic remineralisation and specific regional distribution patterns.