In the NE Atlantic during May 1990, a period of high aggregate abundance, amorphous aggregates contained substantially higher concentrations of bacteria, cyanobacteria and flagellates than the surrounding seawater. Those from 45 to 55 m, at the aggregate maximum just below the seasonal thermocline, contained 2.1 to 25.4 X 10' bacteria, 1.0 to 4.7 X 10' cyanobacteria and 1.3 to 33.0 X 106 flagellates ml-' aggregate. Leucine incorporation rates by bacteria attached to aggregates ranged from 12 to 206 X 10-*' m01 cell-' d-' The concentration of bacteria associated with faecal pellets was generally higher than the concentration on the amorphous aggregates. Bacteria attached to aggregates were equivalent to 10 and 14 % of free-living bacterial carbon integrated through the water column to 100 and 300 m, respectively. This could rise to 25 and 34% if the different carbon content of attached and free-living bacteria was taken into account. However, this study coincided with the maximum marine snow concentration measured 150 km southwest, so at other times when marine snow concentrations are lower, the proportion of attached bacteria will be less. The contribution of attached bacteria to total bacterial production in the top 100 and 300 m ranges between 1.8 and 3.4 %. The contribution of free-living bacterial carbon to suspended POC (particulate organic carbon) was between 25 and 33% and after correcting for their retention on the glass fibre filters, this contribution could be 28 to 40%. Flux studies during 1989 and 1990 indicated that a smaller proportion of POC flux (9 %) and bacterial carbon flux (10%) reached 3100 m than mass flux (25 and 35%), indicating there are processes involved which preferentially utilise or reduce the POC and bacterial components. Bacterial detachment from sinking particles could contribute 2.4 % of the integrated bacterial biomass per annum. The fraction of sink~ng POC lost between 150 and 3100 m may be an important carbon source to the mid/deep-water bacterial population, capable of supplying around 90% of the bacterial carbon demand during April to July 1989 at 47" N, 20°W. In the deeper waters, between 600 and 3100 m, there was sufficient depth-dissipated sinking POC to potentially supply the carbon demand of the bacteria. However, above 600 m an additional source of organic carbon is required to support their growth.