Host plant selection by herbivorous insects involves successive steps of orientation, colonization, probing, and acceptance decisions, which do not always converge. This distinction is relevant for selecting attractive plants, either for trap-cropping strategies or for incorporation into push-pull systems. In this study, we evaluated the potential of curly mustard (Brassica juncea) as a trap plant for the management of Brevicoryne brassicae in kale crops (Brassica oleracea var. acephala), testing whether aphid preference and biological performance are accompanied by olfactory attractiveness. In dual-choice assays, winged aphids preferred mustard plants, which concentrated 70% of the insects. In the performance assay, seven days after infestation with five winged aphids, a significantly higher number of adults and total individuals was observed on mustard than on kale. In a four-arm olfactometer, winged aphids preferred the odors of both plant species over clean air, but preferred kale odors over mustard odors. Volatile profiles differed between the plant species: α-thujene and β-phellandrene were detected in kale, whereas ocimene was identified in mustard. The results reveal a dissociation among the steps of host plant selection, indicating that olfactory attractiveness alone does not determine the colonization and retention of B. brassicae on curly mustard plants. Nevertheless, mustard showed potential as a trap plant and as a pull component in push-pull systems for the management of B. brassicae. These findings indicate that trap plant selection should not be based solely on olfactory attractiveness, but should also integrate retention and population growth, providing more robust criteria for integrated pest management.