The results of sequencing the collared flycatcher genome, and re-sequencing population samples from this species and its sister species, the pied flycatcher, reveal the existence of areas of high sequence divergence compared to background levels, and suggest that complex repeat structures may drive species divergence and that sex chromosomes and autosomes are at different stages of speciation. Flycatchers are important models for speciation. To provide genome-wide insight into the divergence that occurred between lineages during speciation, Hans Ellegren et al. have sequenced the pied flycatcher genome and re-sequenced population samples from this and another species. The results reveal the existence of 'divergence islands' — with fiftyfold higher sequence divergence than the genomic background — non-randomly distributed across the genome. Natural selection acting in both lineages drives divergence in these regions. The authors also report the unexpected observation that targets for selection may not be genes, and may instead be centromeric and telomeric repeats. Unravelling the genomic landscape of divergence between lineages is key to understanding speciation1. The naturally hybridizing collared flycatcher and pied flycatcher are important avian speciation models2,3,4,5,6,7 that show pre- as well as postzygotic isolation8,9. We sequenced and assembled the 1.1-Gb flycatcher genome, physically mapped the assembly to chromosomes using a low-density linkage map10 and re-sequenced population samples of each species. Here we show that the genomic landscape of species differentiation is highly heterogeneous with approximately 50 ‘divergence islands’ showing up to 50-fold higher sequence divergence than the genomic background. These non-randomly distributed islands, with between one and three regions of elevated divergence per chromosome irrespective of chromosome size, are characterized by reduced levels of nucleotide diversity, skewed allele-frequency spectra, elevated levels of linkage disequilibrium and reduced proportions of shared polymorphisms in both species, indicative of parallel episodes of selection. Proximity of divergence peaks to genomic regions resistant to sequence assembly, potentially including centromeres and telomeres, indicate that complex repeat structures may drive species divergence. A much higher background level of species divergence of the Z chromosome, and a lower proportion of shared polymorphisms, indicate that sex chromosomes and autosomes are at different stages of speciation. This study provides a roadmap to the emerging field of speciation genomics.