A mesoscale feature of the ocean exists on a spatial scale of tens to hundreds of kilometres and a temporal scale of one to three years. The size of micronekton (10 mm to 100 mm), their sustained swimming speed (at least 5 km d-1) and their lifespan (typically one to three years) recommend them for study in relation to mesoscale oceanographic features. Micronekton typically migrate on a diel cycle from deeper waters by day to shallower waters by night. A warm-core eddy in the Tasman Sea, Eddy J, was studied as an example of a sustained mesoscale feature that may influence the distribution of micronekton. Two hypotheses were central to the study: that the community of micronekton was consistently distributed in relation to the spatial and temporal scale of the eddy, and that this distribution was maintained by the movement of the animals. The East Australian Current spawned Eddy Jin early 1979. The eddy was identifiable for about 18 months by an isothermal, isohaline core that was warmer by up to s0c than the surrounding water. The core varied in size with the age of the eddy but was typically 250 km in diameter and 250 min depth. Thermal stratification was produced in the central region of the eddy by a shallow 'summer cap' at the surface and by a cooler eddy, Eddy I, which sank beneath Eddy J early in 1980. A strong current flowed anti-clockwise at the edge of the eddy, attaining maximum speeds of the order of 2 m s-1 at the surface. Drogued buoys remained in eddies for periods of up to five months. The micronekton, comprising cephalopods, crustaceans and fishes, were sampled from the eddy. Samples were collected at night by a midwater trawl to a maximum depth of 540 m. Collections were made on each of five cruises from August 1979 to May 1980. Composition and abundance were recorded for 298 species of micronekton in 145 samples. Nearly half of the fishes and cephalopods were caught in the upper 100 m. Crustaceans were caught about 200 m deeper on average than fishes and cephalopods. In the 145 samples of micronekton collected in the vicinity of the eddy, 44 taxa and 1 792 individuals of cephalopods were caught. Relative to crustaceans and fishes, no species of cephalopod was very abundant. Among the four most abundant cephalopod species, differences in the size or abundance of individuals were found between samples from the centre, edge and outside of the eddy but the differences were not consistent from cruise to cruise. The most abundant cephalopod species occurred in similar numbers in the centre, edge and outside of the eddy. On one cruise, large numbers of larvae of this species were found at the edge of the eddy. Multivariate analysis of the cephalopod data revealed. iittle pattern. Samples of cephalopods collected on the same night within a spatial scale of about 10 km were similar each to the other, irrespective of their location relative to the eddy. More species and more individuals of cephalopods were found in near-surface than deeper waters, but depth-relateu differences were not substantial. The number of individuals in each species, or in the cephalopods as a group, was insufficient to reveal pattern on the spatial and temporal scale that was sampled. Species of crustaceans and fishes were more abundant than cephalopods. In the 145 samples, the cru􀀆tac.t.a.-scomprised 54 species and 21 916 individuals, the -tis\.-\e..s WO species and 21 635 individuals. The analysis of all 298 species of micronektcn, comprising 45 343 individuals, was more fruitful than the analysis of cephalopods alone. For an holistic interpretation of the micronekton community in relation to the eddy, each species was treated in the multivariate analysis as an equal attribute of the samples. In practice the more abundant, more variable species were favoured as the most likely contributors to the connnunity pattern. These species were generally crustaceans and fishes. Classification repeatedly grouped together samples of micronekton from the core of Eddy J over the entire ten months that the eddy was sampled. These results, from the analysis of five cruises together, were confirmed when each cruise was analysed separately. The temporal stability within the eddy's core had no counterpart outside the eddy. Stratification of the micronekton within the core was observed at a depth and time that corresponded to the appearance of Eddy I below Eddy J. Despite this addition, the group of samples from the core of Eddy J remained distinct. The group from the Eddy J core was characterized by fishes and cephalopods; the group from the deeper Eddy I core by crustaceans. The grouping of other samples was related to the difference between the first three and the last two cruises, between the warm shallow and the cold deep waters, and between the locations inside and outside the eddy. The temporal stability of the group of micronekton species within Eddy J was remarkable in view of the dynamic physical environment of the eddy and of the mobility of the animals. An animal swimming horizontally from the centre could swim beyond the ecdy in 6 to 125 days. Even the slowest rate was well within the lifespan of the animals and of the eddy itself. The micronekton possessed the mobility to escape the eddy and yet a group of them remained faithful to the core of Eddy J for ten months. This was twice as long as a buoy had remained in an eddy. The inference was drawn that the position of the micronekton in the eddy was being actively maintained by the micronekton themselves. The central question of the orientation capacity remains unresolved: whether micronekton can orient themselves in relation to an eddy. If they cannot, then their swimming ability is irrelevant to their ability to escape from the eddy. If they can, then over what spatial range can they orient themselves to a mesoscale feature? Small tags with transmitters may provide answers to these questions: the technology is available but the expense is formidable. Micronekton are ideal subjects on which to study the coupling of physical and biological mesoscale phenomena: they repeatedly associate with mesoscale features but lack the long-range migrations of the larger pelagic species such as squid and tuna. Increasing our understanding of the causes and mechanisms of association between the micronekton and mesoscale features of the ocean will remain a fruitful area of research in biological oceanography.