Were-emphasize thatwehave not claimed nor have we implied that laboratory data should be used by themselves to characterize the risks of Cry toxins an PIs to natural enemies. We are concerned that only limited inferences can be drawn by a risk assessment process that relies only on laboratory data to make critical initial decisions about risk. We believe that sound generalizations from the laboratory data concerning the responses to genetically modified plants by natural enemies will emerge as this literature is explored in detail. Just as there are not 10 million different types of population dynamics (Lawton 1992), there are not 10 million different types of responses to GM plants. In our paper, we pointed out that these laboratory data paint a more complex picture than hypothesized by Romeis et al. (2006b) and Shelton et al. (2009). The criticisms of Shelton et al. (2009) of our statistical methods do not stand up to scrutiny and do not invalidate one of our main conclusions: existing Cry toxins and PIs have nonzero effects on natural enemies in the laboratory that need to be understood better. We encourage the reader to critically examine their claims in light of the evidence and explanations given in our original paper (Lövei et al. 2009) and the additional clarifications presented here. To close on a positive vein, we note several important findings in our paper (Lövei et al. 2009) that were not disputed by Shelton et al. (2009). These include that (1) the data support only limited generalization about the responses of natural enemies to Cry toxins and PIs; (2) there is an overemphasis on five natural enemy species, although the literature is expanding in scope; (3) tests have been conducted in only a few countries; (4) the Cry toxins that have been studied are mainly Cry1Ab, Cry1Ac, and to a lesser extent Cry3Bb, and other commercialized Cry toxins are under- reported. In addition, we found (Lövei et al. 2009) that (5) parasitoids may be more sensitive than predators to the effects of both Cry toxins and PIs, (6) PIs seem to affect natural enemies more than Cry toxins, and (7) Cry toxins and PIs can have beneficial effects on natural enemies.