Studies of induced plant resistance usually either examine physiological/chemical mechanisms or explore the ecological and evolutionary role of induced resistance. To connect these two methods of study, data are needed that address the relationships between plant chemistry and effects of induced resistance on herbivores under field conditions. In this paper we combine the results of a greenhouse experiment and a field experiment to try to make such a connection. Levels of induced and constitutive resistance to Mexican bean beetles in several soybean genotypes were measured in a greenhouse experiment using a behavioral bioassay. In a field experiment, beetle performance and induced and constitutive levels of cystein proteinase inhibitor activity were measured for these same genotypes. Greenhouse bioassay ratings of induced and constitutive resistance and induced and constitutive levels of proteinase inhibitor activity agree closely for individual genotypes, suggesting that proteinase inhibitors (PIs) are involved in induced resistance. However, while greenhouse bioassay measures of constitutive and induced resistance were good predictors of beetle performance in the field, proteinase inhibitor activity did not predict beetle performance well. While PIs appear to be involved in induced resistance, they were not strongly correlated to Mexican bean beetle performance in the field in this study. These results suggest that measurement of single aspects of plant chemistry may not encompass all ecologically important aspects of plant resistance.