Genome variation during the often-complex speciation process is affected by numerous factors, and can affect the likelihood of evolutionary diversification. Disentangling the processes driving patterns of genome variation is challenging, because different processes can generate similar patterns. In this talk, I will combine data from whole genome re-sequencing of natural and experimentally transplanted populations, genome wide association mapping, and theoretical modeling to test the processes driving genome divergence during ecological speciation in Timema stick insects. The results indicate that even the early stages of speciation can involve numerous genomic regions affected, either directly or indirectly, by divergent natural selection. However, major work remains to be done to objectively determine the traits and genomic regions directly and causally subject to selection and which are most critical for driving speciation. I will discuss how progress on this front can be made using emerging 'ecological' model systems in which genome variation can be experimentally studied in the wild.