CCAMLR Conservation Measure 32/X sets a 1.5 million tonne precautionary catch limit on krill (Euphausia superba) in Statistical Area 48. The measure also implies an application in future of precautionary limits to subareas or local areas of this area. Nine alternative methods of determining subarea or local area krill catch limits are evaluated relative to six criteria: (i) the degree to which information on biological relationships is considered; (ii) the cost of data collection; (iii) the reliability of required information; (iv) the ease of enforcement; (v) the effects on current fishing patterns; and (vi) the potential for delay in implementing the alternative. An alternative is less likely to adversely impact dependent species (e.g., penguins and seals) if the ecological relationships between krill and their predators are explicitly considered and the potential for delayed implementation is low. Therefore, we consider the following trade-off to be important: choosing a biologically explicit alternative and delaying implementation, or choosing a biologically unrealistic alternative and implementing a management scheme immediately. We recognise that other trade-offs may be equally important. Alternatives that allocate the 1.5 million tonne limit by evenly dividing the catch among subareas or by using historical catches to set limits can be categorised as having a low potential for delaying implementation, but they ignore information on biological relationships. Alternatives based on protective zones, critical periods, predator censuses, and predator-prey models include large amounts of biological information, but may not be practical in the near future. Alternatives based on continental shelf area, simple pulse fishing, and krill surveys are not biologically explicit and result in delayed implementation. None of the alternatives are categorised as being both biologically explicit and immediately available for implementation. However, two of the alternatives (i.e., protective zones and critical periods) are unsatisfactory only because they would alter current fishing patterns. These two alternatives could be implemented immediately if the CCAMLR Member nations are willing to tolerate changes in current fishing patterns.