Hybridization is a powerful mechanism to generate biodiversity, for instance through species formation. However, well-supported cases of homoploid hybrid speciation limited to plants and animals, which indicate that either unicellular eukaryotes have been overlooked or that this mechanism is restricted to multicellular organisms. Experiments in the laboratory have shown that new yeast species can be formed by hybridization, suggesting that this mechanism could be, in principle, observed in nature. Using population genomics, experimental crosses and fitness assays, we show that homoploid hybrid speciation took place recently in natural populations of Saccharomyces paradoxus inhabiting the North American temperate forests. We find (1) reproductive isolation between hybrid and parental species, (2) evidence of hybridization in the genome, and (3) a link between hybridization and reproductive isolation. In addition, we found that the hybrid lineage displays specific growth phenotypes that may reflect a new ecological niche. Our results show that hybridization may play a key role in species formation in fungi. Because there is very little opportunity for pre-zygotic isolation among yeast species, this mechanism of species formation could be common in the wild and contribute to fungal biodiversity. Our system will prove useful for the study of the different roles of hybridization, chromosomal rearrangements and hybrid adaptation in speciation.