In 2003, studies revealed that the waters in the lower reaches of Hood Canal in Washington State had very low dissolved-oxygen concentrations, low enough to cause some fish kills between June and October of that year. In order to determine the transport patterns and the persistence of the low oxygen level in this portion of the canal, the U. S. Geological Survey deployed two instrumented platforms on the seabed near the head of the canal that measured currents over the whole water column, water level, near-bed temperature, salinity and oxygen for 2 months in the fall of 2004. Tidal currents, the dominant current component in the canal, flowed primarily along the canal axis and had speeds of 15-20 cm/s. There was also a persistent internal seiche that caused currents to flow along the canal axis with speeds of a few cm/s. The seiche, which had a period of a few days, caused currents in the surface layer to flow in an opposite direction to currents in water depths deeper than 15 m. A pool of warmer, saltier and more oxygenated water moved past the measurement sites toward the head of the canal with a speed of 1 cm/s. CTD measurements taken near the 2 measurement sites during the deployment indicated that this more oxygenated layer of water extended from the bed to the thermocline. Oxygen data from the tripods showed that this water remained in the region until at least the end of October 2004, when the tripods were recovered.