The insecticidal crystal (Cry) proteins produced by the soil bacterium Bacillus thuringiensis have been used for decades to control herbivorous pests. The ability to produce these toxins in transgenic crop plants has given rise to more extensive application in recent years. Little has been accomplished in elucidating the mode of action of Cry toxins, and even less in determining how pests generate resistance to them. One reason for this is that agricultural pests are generally not amenable to molecular genetic analysis. That C. elegans is susceptible to a subset of the toxins in this family makes it a promising system in which to approach these questions genetically. We present findings based on two toxins, Cry5B and Cry21. These toxins are similar in their amino acid sequence (42% identical) and in their destructive effects on the worm gut. Previously, our laboratory has characterized the susceptibility of C. elegans to Cry5B, and isolated mutations in five complementation groups that confer resistance (see abstract by Marroquin et al). Here, we discuss the cloning of one of these resistance genes, bre-5. Mutations in the bre-5 gene result in strong resistance to Cry5B toxin. We have mapped the bre-5 gene to a small interval on chromosome IV. Based on homology, we have identified a candidate gene in this region. We are currently working to determine whether this gene, or some other nearby gene, is the bre-5 gene. We have started to characterize the toxicity of Cry21. It appears to be a more potent nematicide than Cry5B, inducing remarkable destruction of the gut. Additionally, we have not found animals in large mutagenized populations which escape the toxicity of Cry21, even over a series of dilutions. The results outlined incite fundamental questions about what molecules in the worm mediate the toxicity response and the generation of resistance, and about what element or elements in the two toxins under study account for the aforementioned differences in spite of overall high sequence similarity. The experimental approach we have developed will undoubtedly shed light on these questions, and perhaps on issues of normal gut development and maintenance in C. elegans.