Host specificity is fundamental to any project on weed biological control. No insect or other agent should be introduced into a new region before its host range is known. Host range testing determines whether a biological control agent will attack only the target weed or other species, and whether the agent is safe for field release.

Usually, preliminary observations on host range are made in the field in the country of origin. Also the literature is searched for any published accounts of the host range of the potential agent and close relatives. This information needs to be confirmed by more complete host range tests in the laboratory and/or field before the insect is released into a new environment. This section covers the design of host range tests drawing on our knowledge of the principles of host specificity, and host plant selection by phytophagous insects; the methods for screening fungal pathogens are described in the section on Plant Pathogens as Classical Biological Control Agents.

The theory and practice of host specificity testing have been discussed by Cullen (1988), Harley and Forno (1992), and Zwölfer and Harris (1971). The technical guidelines in support of the FAO code of conduct for the import and release of biological control agents contains a section on tests for host specificity but this document is still in draft form (See Approval to Import Agents).

For other information of relevance to host specificity studies, see - Importance of Taxonomy, Compiling a Plant List, Terrestrial and Aquatic Plant Propagation, Host Specificity Workshop Report (Withers et al.) and Deciding on Release Candidates.