Sexual reproduction is an ancient biological process and in most species, has resulted in the evolution of two distinct sexes; females that are typically categorised as producing relatively large and metabolically costly gametes, and males that produce smaller less costly gametes. Such differences between the sexes can result in discordant selection pressures where, for example, males are selected for a fast mating rate, whereas it is beneficial for females to reproduce less often. Because the sexes share a genome, such discordant selection can create an intralocus sexual conflict, where an allele can be simultaneously beneficial in one sex while being detrimental in the other. Ultimately, the resolution of intralocus sexual conflict occurs via the evolution of sexual dimorphism, which allows each sex to approach its individual fitness optimum. A prime mechanism for the evolution of sexual dimorphism is sex-biased gene expression, where males and females express the shared genome differently to produce distinct phenotypes.Sex-biased gene expression (SBGE) appears to be a common feature of dioecious species and during my PhD candidature, I studied the following aspects of the evolution of SBGE in the Australian vinegar fly Drosophila serrata. 1) A deficit of male-biased X-linked genes has been observed in several other species. I assessed the possibility that such a nonrandom distribution of sex-biased genes in D. serrata is a by-product of dosage compensation rather than selection. I found that both selection and dosage compensation likely play a part. 2) While the rapid evolution of male-biased genes is a strikingly consistent pattern of divergence between species, I asked whether the same patterns occur for divergence among populations within species spanning a latitudinal gradient. I found that the patterns are indeed reflective, and I also discovered that many more genes diverged in males than females. Because a lot of the male-specific divergence did not clinally vary with latitude, as might be expected of spatially variable natural selection, I hypothesised that male-driven divergence might be in response to stronger sexual selection on males, which may occur on a population-specific scale. 3) While comparisons of protein coding sequence between species indicate that positive selection is a possible explanation for the rapid evolution of male-biased genes, I found support for several other explanations in relation to the evolution of gene expression of male-biased genes in D. serrata. First, given that any response to selection is proportional to the magnitude of genetic variance, the accelerated evolution of male-biased genes may be facilitated by my finding that male-biased genes in D. serrata possess significantly more genetic variance. Second, evolutionary change may be inhibited by the fact that genes can be involved in multiple biological processes (pleiotropy). For instance, a mutation that enables the gene product to be better at job A can make it worse at job B, the so called “cost of complexity”. While I found that male-biased genes in D. serrata appear less inhibited by overall pleiotropy, a more specific form of pleiotropy that measures the degree to which a gene performs the same biological process in males and females (between-sex pleiotropy) suggested otherwise. 4) Another explanation for the accelerated evolution of male-biased genes is that they are under stronger selection. Using a mutation-accumulation experiment, which uniquely applied a male-limited sexual selection treatment, I assessed this possibility and also explored whether such sex-specific selection was beneficial to both sexes. Although I did not find any indication that selection was stronger on males than females (this may have been due to how fitness was measured), it was interesting to find that sexual selection strengthened the between-sex genetic correlation for fitness, a suggestion that sexual selection primarily purged sexually discordant, rather than concordant, deleterious mutations.