Standard chemical defense theory has long held that plant/insect interactions are at least in part regulated by the plants secondary metabolome. However, this theory does not address whether it is the type of secondary metabolite, their absolute amount or an interaction that is more important in controlling insect responses. We are utilizing the Arabidopsis thaliana glucosinolate system and its impact on specialist and generalist lepidopteran herbivores to try and tease apart these questions. The glucosinolates are biologically inert and upon tissue hydrolysis, they are exposed to the hydrolytic Myrosinase enzyme complex. This complex then activates the glucosinolates into their toxic form. The glucosinolate system in Arabidopsis has natural variation for chemical type, absolute amount, activation rate and activation type. This allows us to test each aspect of variation independently and concordantly to question which aspects of the variation are evolutionarily more important. Further, this research is recently showing that the plant may have developed the ability to detect insect herbivory through by detecting inappropriate glucosinolate hydrolysis. This opens up the possibility that plants may utilize secondary metabolites both as direct action toxins and indirect action signal transduction metabolites to control its interactions with the environment. I will present data on how we are utilizing natural variation through quantitative trait locus mapping, comparative genomics and metabolomics to understand the different roles that glucosinolates play in controlling plant/insect interactions.