An understanding of the geochemistry of the cosmogenic radioisotope Be-10 (t1/2 = 1.5 My) and the stable isotope Be-9 is necessary for the development of their use in deep-sea sediments as a geochronometer and as a monitor of variations in cosmogenic nuclide production. We examined these isotopes in oceanic regions strongly influenced by input of continental materials; the effects of atmospheric aerosols were studied through examination of the water column chemistry of the Mediterranean (a basin strongly influenced by Saharan dust) while salinity transects through the Amazon and Ganges-Brahmaputra estuaries were used to investigate riverine fluxes to the oceans. These results were used to quantify riverine and eolian fluxes; they indicate that eolian dust is probably the most significant source of Be-9 to the oceans. Changes in Be-10:Be-9 in the marine sedimentary record which are not associated with radioactive decay or variability in Be-10 production can thus result from local or regional changes in dust inputs.