Determination of the potential and actual host range of a natural enemy is crucial before its importation and release for biological control. We studied some of the factors that are important in determining the physiological host range of insect parasitoids attacking lepidopteran hosts. Our experimental system consisted of novel host-parasitoid associations, with two New World pyralid stalk borers, Diatraea saccharalis and D. grandiosella; one Old World crambid borer, Ostrinia nubilalis as hosts; and three Old World microgastrine braconids, Cotesia chilonis, C. sesamiae, and C. flavipes as parasitoids. Experiments on the chronology of encapsulation of the parasitoid progeny by host hemocytes indicated that lepidopteran stemborers that are taxonomically, behaviorally and ecologically very similar differ in their ability to encapsulate a parasitoid species. D. saccharalis encapsulated C. flavipes sometimes, whereas D. grandiosella consistently encapsulated C. sesamiae and C. flavipes. C. chilonis was not encapsulated by either Diatraea host. If encapsulation occurred it did not start until four days after parasitization and continued during the following days. O. nubilalis was an unsuitable host for all three parasitoid species; parasitoid eggs were killed within 24 hours of parasitization. O. nubilalis had nearly twice as many hemocytes present in the hemolymph compared to the Diatraea species. In many of the host-parasitoid combinations, there was an initial increase of hemocyte number soon after parasitization, which was not due to mechanical damage at oviposition. There was no correlation between total numbers of hemocytes present in the host hemolymph and the observed encapsulation levels. By understanding the encapsulation response we may be able to make better predictions about the host range of a parasitoid species before its release as a biological control agent.