Host range expansion in insect herbivores is often thought to be mediated by several factors, principal among them is secondary plant chemistry. In weed biological control, the host range of a prospective new agent is one of the most important considerations in the development of control agents. A host testing process seeks to determine the behavioral acceptance and nutritional value of different test plant species to the potential agent. A list of test plants is constructed that comprises species that are close taxonomic relatives of the target weed plus other species of economic or ecologic importance. The host testing of the Melaleuca quinquenervia biological control agent Oxyops vitiosa indicated that larvae would accept and complete development on the Australian target weed Melaleuca quinquenervia, two Australian ornamental species, Callistemon citrina, C. viminalis (all Myrtaceae). Volatile chemistry in these plants functions as behavioral cues used by Oxyops that narrows the host range and prevents damage to non-target plants.