Food is important! Food availability is a major factor influencing reproduction because of the high nutritional demand of gamete production and offspring rearing. Most food resources are spatially and temporally heterogeneous requiring finely tuned sensory perception for optimal foraging. The same holds true for the detection and evaluation of potential mates. The evolution of different sensory abilities, favoured by different environments but also affecting sexual communication, is the basis of the sensory drive hypothesis. This theory provides a potential mechanism for the evolution of pre-mating isolation between populations exploiting different food resources or habitats. To date, evidence for this evolutionary mechanism comes mostly from visual communication in aquatic species living in different light environments. Here, I propose to combine our unique expertise in evolutionary ecology and neurogenetics to explore the role of sensory divergence in reducing gene flow between populations that exploit alternative food resources in Drosophila melanogaster. I will characterize phenotypic and genetic differences of natural D. melanogaster populations (‘ecotypes’) exploiting alternative food resources. Using transcriptomics, I will identify the chemosensory receptors underlying the sensory response of different ecotypes to their food habitats and causally link them to changes in habitat and mate preference. Through a reciprocal transplant experiment, I will establish the contributions of genetic adaptation and phenotypic plasticity to viability, food preference and mate choice. Finally, I will assess whether differences between ecotypes actually translate into gene flow reduction under natural reproductive conditions, including multiple matings and different food environments. Thus, by integrating proximate and ultimate aspects of food-associated population divergence, we will advance our understanding of the role of sensory divergence in reducing gene flow between populations at the single gene level, providing much needed candidate genes and neuronal mechanisms for adaptation and speciation.