An indirect biogeochemical consequence of the eutrophication of shallow coastal marine ecosystems was examined. Changes in manganese distribution in water and bottom sediments, and the flux rates between these two, were followed in a nutrient enrichment experiment using marine mesocosms. Conducted at the Marine Ecosystems Research Laboratory (MERL), Narrangansett, RI, U.S.A., the experiment involved a geometric series of daily nutrient additions (N, P, Si) to six mesocosm tanks (plus three controls) for a 2·3-year period. Changes occurred in sediment-water flux rates of dissolved Mn, in total Mn in the water, and in Mn concentrations of sediment pore waters and solids; the direction and magnitude of change in each of these parameters varied somewhat by treatment. Patterns of changes with increasing levels of nutrient loading were observed, but usually these held only across a limited range of loading or for short periods of time. For example, during the first year the benthic flux of Mn increased from controls to mid-levels of the eutrophication gradient. Moreover, the specific nature of the alteration in Mn cycling varied substantially among treatments, relating to the biological, as well as geochemical conditions, specific to any given tank. For example, a change in the benthic community involving the presence of the benthic shrimp Crangon correlated, qualitatively, with a lowered total water column Mn over the latter months of the experiment, and created a pattern seemingly independent of loading level. Seasonal patterns in Mn concentrations and fluxes, normally evident as correlations with temperature, were occasionally altered; disruption of sediment-water interactions of Mn, which shape the water column concentration of Mn during warm seasons, was significant. Related to this, a general model of benthic Mn flux, as influenced by primary production and eutrophication, as well as the past history of the sediments, was suggested by the data. The rates of dissolved Mn release from sediments, measured in a time series of short-term incubations with an in situ chamber, did not reflect well the longer-term observed changes in sediment in ventory of Mn across the gradient. Progressively larger loss of sediment Mn inventory with increased nutrient loading was suggested; loss generally led to elevated Mn in the water, a change itself determined in each situation by imbalances between the processes of Mn particle deposition and remobilization of Mn 2+ from the sediments. Thus, while the details of Mn cycling varied at different levels of eutrophication and were influenced by the composition and activity of the biological community, the major general consequence of nutrient enrichment was the potential for increased export of Mn from more enriched systems to their offshore receiving waters. Our budget calculations suggested that loss of Mn previously stored within bottom sediments could increase Mn export rates by about 33 to 100% within the first few years of enhanced nutrient loading of a shallow coastal area like Narragansett Bay.