Infectious disease represents an emerging threat to natural populations, particularly when hosts are more susceptible to novel parasites (allopatric) than to parasites from the local area (sympatric). This pattern could arise through evolutionary processes (host populations become adapted to their local parasites and genetically differentiated from other populations at immune-related loci) and/or through ecological interactions (host individuals develop resistance to local parasites through previous exposure). The relative importance of these candidate mechanisms remains unclear. In jawed vertebrates, genes of the major histocompatibility complex (MHC) play a fundamental role in immunity and are compelling candidates for spatially varying selection. We recently showed that song sparrows (Melospiza melodia) are more susceptible to allopatric than to sympatric strains of malaria (Plasmodium). In the current study, to determine whether population differences at MHC explain this pattern, we characterized the peptide-binding regions of MHC (classes I and II) of birds that did or did not become infected in the previous experiment. We recovered up to 4 alleles per individual at class I, implying at least 2 loci, and up to 26 alleles per individual at class II, implying at least 13 loci. Individuals with more class I alleles were less likely to become infected by Plasmodium, consistent with parasite-mediated balancing selection. However, we found no evidence for population genetic differentiation at either class of MHC, based on 36 individuals sequenced. Resistance to sympatric parasites previously described for this system likely stems from individuals’ prior immune experience, not from population differentiation and locally protective alleles at MHC.