Spatial gradients of silver concentrations in the surface waters of San Francisco Bay reveal substantial anthropogenic perturbations of the biogeochemical cycle of the element throughout the estuarine system. The most pronounced perturbations are in the south bay, where dissolved (<0.45 mu m) silver concentrations are as high as 250 pM. This is more than one order-of-magnitude above baseline concentrations in the northern reach of the estuary (6 pM) and approximately two orders-of-magnitude above natural concentrations in adjacent coastal waters (3 pM). The excess silver is primarily attributed to wastewater discharges of industrial silver to the estuary on the order of 20 kg d(-1). The contamination is most evident in the south bay, where wastewater discharges of silver are on the order of 10 kg d(-1) and natural freshwater discharges are relatively insignificant. The limited amount of freshwater flushing in the south bay was exacerbated by persistent drought conditions during the study period. This extended the hydraulic residence time in the south bay (greater than or equal to 160 d), and revealed the apparent seasonal benthic fluxes of silver from anthropogenically contaminated sediments. These were conservatively estimated to average approximate to 16 nmol m(-2) d(-1) in the south bay, which is sufficient to replace all of the dissolved silver in the south bay within 22 d. Benthic fluxes of silver throughout the estuary were estimated to average approximate to 11 nmol m(-2) d(-1), with an annual input of approximately 540 kg yr(-1) of silver to the system. This dwarfs the annual fluvial input of silver during the study period (12 kg yr(-1)), and is equivalent to approximately 10% of the annual anthropogenic input of silver to the estuary (3,700-7,200 kg yr(-1)). It is further speculated that benthic fluxes of silver may be greater than or equal to waste water fluxes of silver during periods of intense diagenic remobilization. However, all inputs of dissolved silver to the estuary are efficiently sorbed by suspended particulates, as evidenced by the relatively constant conditional distribution coefficient for silver throughout the estuary (K-d approximate to 10(5)).