Between March 2000 and April 2001 two commercial fishing vessels undertook trials at South Georgia of a method of fishing for toothfish (Dissostichus eleginoides) using pots. A significant bycatch of lithodid crabs (three species of Paralomis spp.) was encountered. Paralomis spinosissima occurred in shallow water, generally shallower than 700m. P. anamerae, not previously reported from this area, had an intermediate depth distribution from 400 – 800m. P. formosa was present in shallow waters but reached much higher catch levels (and, presumably, densities) between 800 and 1400m. Differences were also noted in depth distribution of the sexes and the size of crabs. Depth, soak time and area were found to significantly influence crab catch rates. Very few crabs (3% of P. spinosissima and 7% of P. formosa) were males above the legal size limit and could therefore be retained. All other crabs were discarded. Most crabs (>99% of P. formosa, >97% of P. spinosissima and >90% of P. anamerae) were lively on arrival on deck and at subsequent discard. Mortality rates estimated from re-immersion experiments indicated that on the vessel which emptied pots directly onto the factory conveyor belt 85% – 90% of crabs would survive discarding, whereas on the vessel where crabs were emptied down a vertical chute prior to being sorted, survivorship was 39% - 58%. P. anamerae was the most vulnerable of the three species to handling onboard and subsequent discarding. P. spinosissima seemed to be more vulnerable than P. formosa.