The pre-release risk assessment, a recommended practice in biological control programmes, was carried out before introducing Closterocerus chamaeleon (Girault), a eulophid parasitoid of the eucalypt gall maker Ophelimus maskelli (Ashmead) (Hymenoptera, Eulophidae), in Sicily, Italy. We evaluated its host specificity in laboratory small arena no-choice tests on six non-target hosts, using O. maskelli as a control species. The non-target species fit with at least one of the following criteria: a) common geographical origin; b) ecological or behavioural affinities with the target host; c) concealed habit of the preimaginal stages; d) taxonomic affinity; e) taxonomic affinity with hosts of congeneric parasitoids of the candidate agent; and f) presence in Italy/Sicily. Closterocerus chamaeleon performed an oviposition sequence of 10 recognisable behaviours, always leading to parasitization when exposed to O. maskelli. The same behaviour sequence was carried out on the gall maker Leptocybe invasa Fisher et La Salle, the only non-target species living on Eucalyptus, but just until Step 6, i.e. until ovipositor extraction and insertion, and never until egg-laying. All of the other non-target hosts were ignored. Furthermore, the adult parasitoid longevity was assessed in three rearing conditions: 1) no food/no parasitisation permitted; 2) food/no parasitisation permitted; and 3) food provided after parasitisation. Adult wasp longevity was significantly higher in treatment 2 (21.1 days ± 4.1 SE) than in treatment 1 (3.6 days ± 0.5 SE) or 3 (3.8 days ± 0.6 SE). Laboratory tests carried out in this study on C. chamaeleon indicated high host selection ability, later confirmed by our post-release field data