Abstract
The “new approach” to selecting biological control agents, as proposed by Hokkanen and Pimentel in 1984, through the use of new exploiter–victim associations is inappropriate for the biological control of weeds because of certain misconceptions in their method. Examples of biological control of weeds cited in their analysis were biased towards cactaceous insects, and cacti are not representative of target weeds. Several of their “new” associations were inaccurate. These inaccuracies are discussed and additional examples are provided to refute the proposed method. Contrary to the recommendation of Hokkanen and Pimentel, new exploiter-victim associations offer limited opportunities for biological control of non-cactaceous weeds with insects, and should not be used as the preferred method in selecting biotic agents for the biological control of weeds.