We investigated the composition, recycling, and mass accumulation rates of sediments along a transect in the Southern Ocean located from 66 degree S to 57 degree S at 170 degree W. This transect also corresponds to the location of a sediment trap mooring line. The sediments at the seven sites studied range from largely terrigenous material to nearly pure (> 90%) biogenic silica. CaCO sub(3) is a minor but persistent component at most sites. Mass accumulation rates have been determined on the basis of excess super(230)Th in the sediments, i.e., super(230)Th-normalized accumulation rates. The influence of redistribution of sediments on the sea floor has been estimated from super(14)C analyses. The recycling of material delivered to the sediments has been characterized on the basis of pore water studies that make extensive use of both in situ sampling and shipboard extractions. The influence of the highly variable rates of input of particulate matter that characterize much of the Southern Ocean upon pore water gradients and fluxes across the sediment water interface has been considered. We find only poor correspondence between BSiO sub(2) burial fraction (=burial/particulate flux), a quantifiable measure of preservation efficiency, and BSiO sub(2) particulate rain along the transect. However, preservation does appear to be closely linked to a combination of sedimentation rate and particulate rain. The burial fraction of BSiO sub(2) is small relative to benthic rain (5-19%). Despite the small fraction buried, burial flux normalized to (sedimentation rate) super(1/2) appears to provide a very consistent means of predicting benthic particulate rain over a large range of rain rates, including data from a number of different studies and environments. At sites with BSiO sub(2) rain greater than or equal to 250 mmol m super(-2) yr super(-1) the average difference between predicted and observed rain is 25-30%. Such rain rates occur in many marine areas, particularly the Southern Ocean, with the result that this relationship potentially provides a means of estimating BSiO sub(2) benthic rain over prolonged periods in the past on the basis of readily measured sediment parameters.