Host plant shift is a major diversification process in herbivorous insects, which constitute one fourth of the Earth’s biodiversity. However, the underlying genetic and evolutionary mechanisms remain unclear. Of particular interest are: (1) the degree of genomic modularity controlling different fitness attributes on the new host, (2) the genetic linkage between host use and reproductive isolation traits, and (3) the repeatability of genetic changes underlying the convergent specialization on the same host. In spite of significant progress, the lack of powerful genetic and genomic tools capable to identify genes underlying host plant use and/or reproductive isolation in classical herbivorous insect models has hindered the resolution of these questions. Here, we propose to investigate an interesting case, where two species, Drosophila sechellia and D. yakuba mayottensis have independently become specialists on the toxic fruits of noni (Morinda citrifolia) in the Seychelles and Mayotte islands, respectively. In both cases, noni specialization was accompanied by partial reproductive isolation. We will combine precise phenotypic analyses with population and quantitative genomics in D. yakuba to identify genes potentially underlying noni use and reproductive isolation in D. y. mayottensis. We will then leverage advanced genome editing and transgenesis tools (e.g., CRISPR\Cas9) that have recently been developed in both species, to functionally dissect those candidate genes. The expected results will improve our understanding of the genetic and evolutionary mechanisms underlying convergent host shift and ecological speciation, which could go beyond the Drosophila-noni relationship towards the identification of neuronal circuits or detoxification pathways that are common among herbivorous insects.