During vertebrate evolution, complex organs have evolved several times. The development of these organs is encoded in the genome. However, it is currently unclear how new complex organs, consisting of multiple interlocking parts, can evolve as a result of genomic change. In this thesis, I aim to find the genomic basis of one such newly evolved organ: the placenta in the livebearing fish family Poeciliidae. In this family, a placenta has evolved nine times independently, allowing for the investigation of multiple evolutionary origins of the same organ within a single group of species. First, I sequence and assemble the genomes of both placental and non-placental poeciliid species. Then, I compare the genomes of placental species with the genomes of non-placental species, aiming to find consistent genomic differences between placental and non-placental species that can be associated with placenta evolution. I show that indeed, placental species show consistent mutations in both protein-coding and regulatory regions of the genome. Protein-coding mutations occur mainly around structural and metabolic genes, while regulatory changes occur mainly around developmental genes. I also show that, contrary to some predictions, gene duplications are not associated with placenta evolution in poeciliid fish. Finally, I show that allele-specific DNA methylation is present in the poeciliid fish Poeciliopsis gracilis, and that its inheritance is nonrandom but instead depends on parent-of-origin of the methylated allele, suggesting genomic imprinting. Together I provide a comprehensive overview of genome evolution in the fish family Poeciliidae, and provide new insights into the evolution of complex traits.