In this study we used reciprocal rearing experiments to test the hypothesis that there is a genetic basis for the adaptive differences in host-use traits among host-associated soapberry bug populations (described in Carroll and Boyd 1992). These experiments were conducted on two host races from Florida, in which differences in beak length and development were found between natural populations on a native host plant species and those on a recently introduced plant species (colonized mainly post-1950). Performance was generally superior on the host species from which each lab population originated (i.e., on the "Home" host species): in analysis of variance, there was significant population-by-host interaction for size, development time, and growth rate. These results indicate that the population differences in nature are evolved rather than host induced. Increased performance on the introduced host was accompanied by reduced performance on the native host, a pattern that could theoretically promote further differentiation between the host races.