Many hypotheses to explain the evolution of feeding specialization among insect herbivores assume a trade-off in fitness across host plant species, i.e., that adaptation to one host (or host taxon) necessarily results in loss of adaptation to at least some other hosts (or host taxa). As potentially powerful and intuitively appealing as this "Jack of all trades is master of none" hypothesis may be, we presently know little about how common fitness trade-offs across hosts are, or about which mechanisms might give rise to such fitness trade-offs. In this study, I tested the hypothesis that differences among plant species in secondary chemistry give rise to fitness trade-offs across host plant species, i.e., that adaptation to one type of secondary compound necessarily results in loss of adaptation to other types of secondary compounds.