Research on species formation and adaptive divergence aims at identifying the proximate and ultimate causes of phenotypic differentiation, which involves complex interactions between genes and the environment. We used a reverse ecology approach to investigate the molecular bases of an on-going speciation event in the budding yeast Saccharomyces paradoxus in the North American deciduous forests. We applied whole-genome sequencing to identify candidate genes under positive selection and high-throughput phenotyping to determine the conditions into which these genes may play a role. We confirmed the fitness effect of protein divergence at candidate loci through direct allelic competition assays. This comprehensive approach allowed us to gain insights in the complex network of genes and environmental factors that may have shaped the fitness determinants of these two incipient species over the last 100,000 years.