ABSTRACT Senecio madagascariensis Poiret (Asteraceae) is a target for biological control in Australia, due to substantial impacts on agriculture and native environments. Studies in the plant’s native range in South Africa prioritised the stem-boring weevil Gasteroclisus tricostalis (Thunberg) (Curculionidae) as a candidate agent. In this study, we assessed the biology and host specificity of G. tricostalis during laboratory trials conducted in both South Africa and Australia. No-choice trials in South Africa revealed that, despite oviposition on 15 non-target Senecio species (seven Australian and eight South African), there was significantly lower oviposition relative to S. madagascariensis, with no survival to pupation or adulthood on any non-target species. In contrast, no-choice tests in quarantine in Australia, recorded oviposition on 12 non-target Australian native Senecio species, with oviposition on two species not significantly different to that on S. madagascariensis. Furthermore in Australia, larvae developed to adulthood on five non-target species, in numbers that were not significantly different to S. madagascariensis. These discrepancies can mostly be explained by the lower phylogenetic distance from S. madagascariensis to the susceptible non-target species tested in Australia and, likely, optimal plant growth conditions in quarantine in two cases. Although no-choice tests are renowned for their conservative nature, preliminary paired-choice trials in both South Africa and Australia failed to demonstrate clear oviposition preferences for S. madagascariensis. These results suggest that, given the high diversity of the genus Senecio in Australia, the weevil is likely unsuitable for release and further trials were suspended.