New control strategies for insect pests of arable agriculture are needed to reduce current dependence on synthetic insecticides, the use of which is unsustainable. We investigated the potential of a simple control strategy to protect spring-sown oilseed rape, Brassica napus L. (Brassicaceae), from two major inflorescence pests: the pollen beetle, Meligethes aeneus (Fabricius) (Coleoptera: Nitidulidae), and the seed weevil, Ceutorhynchus assimilis (Paykull) (Coleoptera: Curculionidae), through exploitation of their host plant preferences. The strategy comprised, for the main crop, Starlight [an oilseed rape cultivar with relatively low proportions of alkenyl glucosinolates in the leaves (thereby releasing lower levels of attractive isothiocyanates than conventional cultivars)] and turnip rape, Brassica rapa (L.) (Brassicaceae), as a trap crop. We tested the system in laboratory, polytunnel semifield arena, and field experiments. The odours of Starlight were less attractive in olfactometer tests to both pests than those from a conventional cultivar, Canyon, and the plants were less heavily colonized in both polytunnel and field experiments. Turnip rape showed good potential as a trap crop for oilseed rape pests, particularly the pollen beetle as its odour was more attractive to both pests than that of oilseed rape. Polytunnel and field experiments showed the importance of relative growth stage in the system. As turnip rape flowers earlier than oilseed rape, beetles would be maintained on turnip rape past the damage-susceptible growth stage of oilseed rape. The development of a pest control regime based on this strategy is discussed.