The contemporary view of insect migration is one of a behaviourally distinct form of movement that is an adaptation to temporary habitats. Migratory insects are characterised by a syndrome of behavioural and physiological traits that promote migration. The migration syndromes of particular species are variations on the general syndrome, and depend on both selective pressures from the environment and the evolutionary response that is supported by the genetics of the syndrome traits. The pattern and role of migration in a particular population are determined by the interaction between the environment and the species’ migration syndrome. Complete understanding of migration in a population therefore requires a combined consideration of the population’s environment, migration syndrome and genetics. A recently developed conceptual model of a generalised insect migration system facilitates such a holistic treatment of insect migration. The model is built around four components: the migration arena (the environment); the population trajectory (the changing population demography); the migration syndrome (the traits that implement migration and determine the fitness of migrants); and the genetic complex underlying the migration syndrome. Specific variants of the model can be derived for particular species and locations. The migratory moth species Helicoverpa punctigera (Wallengren) and Helicov-erpa armigera (H¨ ubner) are widely distributed in Australia, and occur in both crop-ping and non-cropping areas. The habitats occupied by the species are often tem-porary and geographically separated. Moths colonise new habitats by long-distance migration. Both species are abundant in the periphery of the continent during summer. In winter, H. punctigera is almost exclusively confined to the far inland, whereas H. armigera mostly remains near the coast. This thesis describes the development and application of conceptual models of the migration systems of H. punctigera and H. armigera in Australia. The model for H. punctigera was used to formulate and test hypotheses on the species’ population trajectory. That for H. armigera was used to determine the effects of differences between the migration syndromes of the two species on the trajectory. The models incorporated data from empirical models and simulation models developed for selected parts of the migration systems. Bioclimatic models estimated the portion of the continent that is climatically favourable for colonisation by each species in each season. A second type of bioclimatic model estimated the effect of year-to-year variation in habitat quality on the distribution of H. punctigera breeding in the inland. Finally, a simulation model of moth migration estimated the frequency with which wind patterns support various migration pathways. For H. punctigera the portion of the continent predicted to be climatically favourable for breeding tended to cycle between the inland and the periphery, al-though, contrary to observations, central Queensland was predicted to be favourable year-round. In the far inland the winter breeding period was predicted to be sub-stantially shorter in Western Australia than in south-western Queensland. For H. armigera the potential distribution tended to expand and contract around a core area in central Queensland. The probability of H. punctigera breeding at a given site in the far inland during winter was positively correlated with the increase in vegetation greenness at the site between April and July. It was also related to soil and climate variables. Although breeding was predicted to occur more often in some areas than in others, the esti-mated distribution of breeding varied greatly from year to year, and breeding was predicted at any given site in only a minority of years. Validation analyses indicated that the migration model was sufficiently accu-rate for use in characterising the migration system. However, the limited vertical structure present in the available weather data (e.g. representing low-level jets) will prevent the model from confidently forecasting particular migration events. Complete annual cycles of the population trajectory of H. punctigera could be constructed within the limits of the migration arena and the known migration syn-drome. In eastern Australia the cycle included return migration from the eastern cropping areas to the far inland during autumn, as has previously been proposed in the literature. In Western Australia two trajectories were possible—one including spring diapause and a late summer generation in the south-west, and one including a summer generation in the far south-west. Additional trajectories were possible when pupae in spring diapause were assumed to emerge later than has been observed, or when summer quiescence was assumed to be more frequent than would be expected from our currently limited knowledge of its induction. These assumptions enabled H. punctigera to persist throughout the year in the far inland of eastern Australia, but not in that of the west. The conceptual models enabled certain hypotheses on the population trajectories to be tested with existing field and biological data. For other hypotheses they identified data that would enable testing. For a decision on which, if any, of the above H. punctigera trajectories operate in nature, we need more information on the frequency and duration of spring diapause and summer quiescence, plus additional data on breeding activity in parts of Western Australia.