The hydrography of surface waters in and near the Amazon River in February and March of 1973 was indicative of "flood" conditions, both in terms of the offshore extension of the freshwater plume and the high concentrations of suspended matter within the river water. While the net transport of Amazon water was NW (due to the Guiana Current), seasonal changes in wind direction resulted in formation of a nearshore counter current along the Para coast. More than 95% of the terrigenous sediment within the Amazon surface waters settles out in the river mouth, before salinity reaches 3 per thousand . Most of this sediment is resuspended by tides, waves and currents; practically no terrigenous sediment appears to escape offshore. The prime components of suspended matter in the brackish water plume off the Amazon are diatom frustules. Contrary to the suggestions of other workers, Amazon outflow markedly increases the fertility of shelf waters. However, the high production of diatoms is not reflected in the sediments, probably due to rapid recycling as well as to NW transport of frustules by the Guiana Current. Although Amazon-borne silica is initially removed by biologic means, recycling releases it; the ultimate fate of the silica is unknown.