Red imported fire ant (RIFA), Solenopsis invicta Buren, which is an invasive alien ant, causes serious ecological and public safety problems. Concerns about the impact of traditional synthetic pesticides on nontarget organisms, plant essential oils (EOs) as alternative products with lower environmental impact can be used to control insects. The "push-pull" strategy is considered to be a method of controlling the number and distribution of pests through attraction and repulsion. To find active compounds with effect on the behavior of RIFA, 6 types of EO of plant flowers with a strong aroma, including Osmanthus fragrans (Thunb.) Lour., Sophora japonica Linn, Eugenia caryophμllata Thunb., Ligustrum compactum (Wall. exG. Don) Hook.f., Jasminum sambac (L.) Ait., Allium polyrhizum Turcz. Ex Regel were studied by Y-tube experiment, electrophysiological detection (EAG), and gas chromatography-electroantennographic detection (GC-EAD). The results showed that different concentrations of six EOs significantly affected the behavioral selectivity of different castes of RIFA, the data of high concentration-repellent and low concentration-attractive have been recorded. The repellent rate of 1000 μg/ml L. compactum of EO was 68.75 % and the attraction rate of 2 μg/ml of EO was 66.25 % on small worker ants. For large worker ants, both 1000 μg/ml (repellent rate: 72.50 %) and 40 μg/ml (attractive rate: 68.75 %) of L. compactum EO reached an extremely significant difference. As for E. caryophμllata EOs, the repellent rate of male ants was 67.5 % at 1000 μg/ml and 68.75 % of the attraction rate was calculated at 2 μg/ml. The most interesting finding was that all concentrations of L. compactum EOs showed a remarkable attractive effect on virgin ants and the attraction rate of 1000 μg/ml EO reached an extremely significant difference (attraction rate: 72.5 %). The highest absolute EAG values of different castes of RIFA in the EO were followed as: small worker ants (1.8788 mV, J. sambac), large worker ants (1.5550 mV, O. fragrans), male ants (1.6088 mV, A. polyrhizum), and virgin ants (1.2350 mV, E. caryophμllata). The GC-MS results of the EOs were followed as: the compounds with high content in O. fragrans EO were butyl phthalate (37.95 %) and p-methoxyphenethyl alcohol (16.76 %); the main chemical components in S. japonica EO were benzeneethanol (20.62 %) and phenol (11.16 %); the total contents of eugenol (82.5 %) and aceteugenol (15.81 %) reached 98.31 % in E. caryophμllata EO; the most compound of EO was benzyl alcohol (19.33 %), followed by linalool (18.70 %) in J. sambac EO; the main chemical components of L. compactum EO were benzyl alcohol (17.45 %) and phenethyl alcohol (27.63 %); five compounds, including pentacosane (11.78 %), heptacosane (10.37 %), oleylamide (11.9 %), non-acosane (20.15 %), and octacosane (8.24 %) were the main chemical components in A. polyrhizum EO. The GC-EAD results showed that three compounds (estragole, α-humulene, and caffeine) were identified to have the potential to develop as behavioral orienting chemicals against worker ants of RIFA. Furthermore, several compounds were expected to be developed as attractants for reproductive ants, including eugenol, cis-linalool oxide, furfural, and its analogs. Our data will provide basic materials for the "push-pull" strategy and have the potential to lay the foundation for the eventual development of efficient and safe repellents.