Many studies have used host-parasite systems to study local adaptation but few of these studies have found unequivocal evidence for adaptation. One potential reason is that most studies have focused on limited measures of host and parasite fitness that are generally assumed to be under negative frequency-dependent selection. We have used reciprocal cross-infection experiments to test for local adaptation in Hawaiian, South Floridian, and eastern North American populations of monarch butterflies and their protozoan parasites. Sympatric host-parasite combinations did not result in greater host or parasite fitness, as would be expected under coevolutionary dynamics driven by negative frequency-dependent selection. Instead, we found that Hawaiian hosts were more resistant and carried more infective and virulent parasites, which is consistent with theoretical predictions for virulence evolution and coevolutionary arms race dynamics. We also found that Hawaiian hosts were more tolerant, particularly of Hawaiian parasites, indicating that increased resistance does not preclude increased tolerance within a population and hosts may be more tolerant of local parasites. We did not find a similar pattern in the South Floridian or eastern populations, possibly because host-parasite adaptation occurs within the context of the greater ecological community.