MHC region genes have been the subject of molecular evolutionary studies both from single species and from a variety of taxa. The African clawed frog, Xenopus laevis, provides a good model for the study of immune genes such as the MHC class Ia because of the genomic architecture of the MHC region. Herein, I investigate 1) the molecular evolution of the MHC class Ia gene at the population level in X laevis, and 2) the evolution of proteasome subunits psmb5 and lmp7 following duplication from their common ancestral locus using a phylogenetic sampling of mainly vertebrate taxa. Modelbased maximum likelihood statistical procedures are used in an effort to overcome typical problems associated with complex patterns of molecular evolution at these loci. In this thesis I present several new findings, and Chapters I and II focus on phylogenetic investigation of proteasome subunits. Results indicate that several evolutionary mechanisms operate on lmp7 that makes phylogenetic reconstruction of this locus difficult. I show that analysis of this gene is sensitive to the particular assumptions of various models of nucleotide evolution commonly used for phylogenetics. I also investigate whether or not natural selection operated differentially on duplicates of the proto-lmp7 gene locus. I provide evidence that positive Darwinian evolution contributed to the functional divergence of gene family members derived from this locus-making this one of the few examples of positive natural selection operating on a protein with housekeeping functions. Several new and major findings are also presented for the X. laevis class la MHC gene in Chapters III, IV and V. For the first time I provide robust estimates of substitution rates that show the operation of natural selection on peptide binding region (PBR) amino acids of the class la gene. I also show for the first time that intralocus recombinations are a major source of variation in the class la gene in X. laevis. Patterns of polymorphism at the class la locus are investigated in greater detail, and provide evidence for a molecular basis driving the coevolution of functionally linked genes. Combining data from other species, my results also demonstrate that the mode of MHC class la evolution is different than the classical paradigm detailed in mammals. Finally, my research is the first to demonstrate that non-linkage of the class I and class II genes in a single genomic region is not always necessary for this mode of class la evolution, as previously expected.