This study examines high rate vermicomposting of sewage sludge using high stocking densities of earthworms. To examine the loading capacity, two vermicompost beds were established in identical 0.84 in diameter reactors, one loaded at an average rate of 10 kg-wet-sludge-mixture m(-2) day(-1) (0.5 kg-carbon m(-2) day(-1)), the other loaded at 20 kg-wet-sludge-mixture m(-2) day(-1) (1 kg-carbon m(-2) day(-1)). The sludge mixture was from a commercial vermicomposting company (Vermitecho) and contained 80-90% sludge and 10-20% green waste and clay. The beds were operated in fed-batch mode for 38 days, and then monitored for a further 12 days without any further sludge loading. Earthworms (Eisenia fetida) were added once or twice weekly over the 38 days loading period to gradually decrease the ratio of sludge loading rate to total earthworm biomass. By adding earthworm incrementally, the feeding rate ranged from 3.9 to 1.25 kg-wet-sludge kg-earthworm(-1) day(-1) for the full load experiment and 2-0.62 kg-wet-sludge kg-earthworm-1 day(-1) for the half load experiment. The extent of degradation was estimated by fitting a 1st order model to the CO2 production rate from the beds. Based on the 1st order model, 53 +/- 20% (95% CI) and 68 +/- 4% of the organic carbon was converted to CO2-C in the full load and half load experiments respectively. The CO2 production rate in the half load experiment became stable and repeatable when the total earthworm biomass reached 5.4 kg, corresponding to a feed rate of 1.04 kg-wet-sludge-mixture kg-earth worm(-1) 1 day(-1). In contrast, the rate of CO2 production was still climbing and traces of methane were evident in the full load experiment at the end of the 38 day loading period. The experiments indicate that high rate vermicomposting beds are sustainable providing the feeding rate does not exceed approximately 1 kg-wet-sludge kg-earthworm(-1) day(-1). (c) 2006 Elsevier Ltd. All rights reserved.