In a 14-d period 38 swarms of Antarctic krill,Euphausia superba Dana, were sampled in an area 55.5 x 55.5 km to the southwest of Elephant Island, at the north of the South Shetland Islands. Moult stage, maturity stage, and size of ca. 100 krill from each swarm were measured. Each of the characteristics varied greatly between swarms. Moulting krill were found in most swarms, but in one swarm all of the krill were just about to moult. Ways in which moulting may act as a possible sorting mechanism are discussed. Data for all the analysed krill (ca. 3000 specimens) were used to investigate the interdependence of moult rate, sexual maturation and growth. While all immature krill moulted at approximately the saine rate in the study, there were significant differences in the moulting rates of mature male and female krill. Gravid female krill continued to moult, although less frequently than mature males. As a consequence males had to attach spermatophores to females after each moult. It is likely that variation in moult rate of females and males was related to the energy expenditure required for ovary development in females and spermatophore production and searching behaviour in males. Spawning and moulting were only partly coupled in gravid females. Spawning appeared to take place predominantly during Moult Stage D2. There was no evidence of intermoult growth by intersegmental dilation.