The model of the krill fishery presented in 1993 (wg-Krill-93/14) refined and applied to both Japanese and Chilean krill fishing fleets in Subarea 48.1. A stochastic element is introduced to account for variability in catch rates. The most successful management regime is found to be one which restricts fishing within 75 km of breeding penguins during January and February. For this regime the model predicts a 90% reduction in overlap with foraging predators, and a 15 to 20% reduction in catch. Closure of Livingston Island for the breeding period resulted in a 60% reduction in overlap with predators and 0 to 15% reduction in catches.