Measurements were made of resource availabihty and the filtration rate and particle retention efficiency of mussels M y t h s edulis L. from an estuarine site on the River Lynher and from an open coastal site at Whitsand Bay, Cornwall, UK. Small cocci of 0.25 km, large cocci of 0.56 pm and small rods of 0.725 pm mean spherical diameter comprised 87 % of the free-living bacterial population at Whitsand Bay and 86% in the Lynher estuary during September 1984. Free-living bacteria represented only 5 % of the carbon and 9.7 and 11 3 O/O respectively of the nitrogen potentially available for exploitation as a food resource in the water column at the estuanne and open coast sites. Weightspecific clearance rate of particles > 12.7 km diameter by M. edulis from the Lynher estuary was approximately 2.5 1 g-' h-' at 14'C whdst that for mussels from Whltsand Bay was 2.6 1 g-' h-'. Relative retention efficiency of particles was generally similar for mussels from the 2 sites, declining to ca 28 O/O for natural bacterioplankton of 0.513 to 0.524 pm mean spherical diameter. These results suggest no significant intraspecific differences in filtration rates or relatlve particle retention efficiencies, for mussels from the 2 sites, in spite of major differences in partlcle concentrations at the sites. Although ca 65 O/O of the natural bacterioplankton were cleared from the expenmental vessels after 335 min at 14 'C, carbon and nitrogen yield from bacterioplankton at natural concentrations is likely to be small. The complete particle spectrum, represented by living phytoplankton cells, detritus and bacterioplankton, would need to be exploited by the Whitsand Bay population of mussels to sustain carbon and nitrogen demands since the total yleld from all particulate size classes was only 713 ug C h-' and 78 pg N h-' compared with an estimated minimum requirement of 737 pg C h-' and 45.5 ug N h-'. The total carbon and nitrogen resource in the Lynher estuary was well in excess of the estimated requirements of the mussels. However, it is estimated that the free-living bacterial resource would contribute only 4.2 % to the carbon budget and 17 % to the nitrogen budget of the Lynher estuary mussels compared with 1.6 and 7 % respectively of that of the open coast mussels. These results suggest that mussels from both the estuarine and coastal sites utilize phytoplankton or similar sized particles as a primary nutritional resource, although in other systems where bacterial standing stocks are high relative to other particulate resources, bacterioplankton may make a more important contribution to the carbon and nitrogen requirements of filter-feedmg organisms.