This paper describes research on the cinnabar moth, Callimorpha jacobaeae, introduced from England and Italy to Australia for the control of the poisonous weed, ragwort (Senecio jacobaea). Studies were conducted for six seasons in a high rainfall area of southern Gippsland, Vic., where the impact of ragwort infestation on dairy pastures was severe. Callimorpha has a univoltine life cycle with an obligatory pupal diapause during winter. The termination of this diapause in introduced stocks was successful to a limited extent only, and changes in its duration required for synchronization with the southern hemisphere seasons adversely affected the reproductive capacity of emerging females. The larval progeny of Italian stock failed to survive in the field, and disappeared completely within the first season, whereas those of English origin were reasonably successful. In view of the low reproductive rate of Callimorpha and because of its numerous insect enemies, breeding was carried out in the field using techniques designed to provide protection from the locally abundant predators. An attempt was also made to assess factors playing a role in its numerical regulation. The larvae were liable to heavy mortality in the field due to the combined effect of fungal infections, insect predators, and parasites, and also to a virus disease introduced in a latent state with the insect. At an early stage of this work (the second Australian generation), a nuclear polyhedral virus epizootic destroyed 90% of the larval population. The fifth Australian generation was the last to survive in the field. Of the larval predators, the mecopteran Harpobittacus nigriceps caused the most serious mortality. This predator, common in all ragwort-infested areas of Victoria, showed a zonal pattern in its density distribution; the high density zones often overlapped sites occupied by larval colonies of the cinnabar moth. The larvae of Callimorpha were more frequently taken by Harpobittacus than tipulid flies (Macromastix spp.) which are normally its principal prey; the abundance and distribution of these flies was also studied. The abundance of both Harpobittacus and Macromastix, was influenced by weather factors, resulting in a marked annual fluctuation in their numbers. At times of high Harpobittacus abundance, mortality in larval colonies of Callimorpha due to this predator averaged over 80%, with extremes of 90-100% being observed frequently. The larvae of Callimorpha were found to be potentially efficient in controlling ragwort. The viable seed production of severely attacked plants was reduced, following defoliation and destruction of the primary flowerheads, by an average of over 98%. However, taking all factors into account, it is unlikely that Callimorpha could exert useful control of ragwort in Australia. There seems little chance that high larval densities, necessary to suppress seeding by ragwort, could be maintained in face of heavy predation. As the data presented in this paper indicate, Harpobittacus alone would be capable of preventing the effective establishment of C. jacobaeae in this country.