This study defines the composition and biomass charactenstics of 5 of 6 previously described faunal assemblages that form a mosaic community on hydrothermal sulfide edifices of the Juan de Fuca Ridge (northeast Pacific). Quantitative samples of each assemblage were acquired during 'ROPOS' Remote Operated Vehicle (ROV) dive programs in 1994 and 1996. Total abundance, and species nchness, as well as wet and dry weights, were calculated for each assemblage, and sampled surface area was measured directly or from scaled images of sample scars. These data were used to compare species composition, nchness and biomass of the distinct assemblages and to estimate total biomass of a sulfide edifice. In addition to major compositional differences. we observed an increase in density, biomass and species richness along a proposed successional sequence from the Paralvinella suifincola (annelid polychaete) assemblage (Assemblage I) to the low-flow Ridgeia piscesae (vestimentifera) community (Assemblage V-LF). Biomass (dry weight without vestimentiferan tubes) of the different sampled assemblages ranged from 0.011 kg m-2 for Assemblage I to 4.68 kg m-2 for Assemblage V-LF and 2.33 kg m-2 for the rarer high-flow Ridgeia pjscesae community (V-HF). Resulting quantitative information was used to refine a previous model of community succession. Comparisons with other manne ecosystems showed that the biomass of these and other hydrothermal assemblages dominated by symbiont-bearing organisms (vestimentifera, bivalvia) is sirnilar to those found in the most productive photosynthetically based assemblages. Tubeworm growth and sulfide accretion greatly increase total surface area available to vent organisms, and may attenuate competition for space. The 3-dimensional habitat formed by Ridgeia piscesae tubes may influence species distribution and enhance species nchness. The tube worm assemblages comprise the major and probably the most stable component of total edifice biomass. At one site, over a 4 yr period, there was substantial environmental change and major shifts in coverage by other assemblages but relatively little change in total coverage by R. piscesae. As a result total edifice biomass (219 to 251 kg dry weight) varied by only 16 % over 1 to 3 yr intervals. Considerable quantitative ecological information can be derived from analyses of submersible-collected imagery, with sampling serving pnmanly as a ground truthing tool. Limitations of sampling and surface area determinations are considered.