Most plant pathogens exhibit host specificity but when former barriers to infection break down, new diseases can rapidly emerge. For a number of fungal diseases, there is increasing evidence that hybridization plays a major role in driving host jumps. However, the relative contributions of existing variation versus new mutations in adapting to new host(s) is unclear. Here we reconstruct the evolutionary history of two recently emerged populations of the fungus Pyricularia oryzae that are responsible for two new plant diseases: wheat blast and grey leaf spot of ryegrasses. We provide evidence that wheat blast/grey leaf spot evolved through two distinct mating episodes: the first occurred ~60 years ago, when a fungal individual adapted to Eleusine mated with another individual from Urochloa. Then, about 10 years later, a single progeny from this cross underwent a series of matings with a small number of individuals from three additional host-specialized populations. These matings introduced non-functional alleles of two key host-specificity factors, whose recombination in a multi-hybrid swarm probably facilitated the host jump. We show that very few mutations have arisen since the founding event and a majority are private to individual isolates. Thus, adaptation to the wheat or Lolium hosts appears to have been instantaneous, and driven entirely by selection on repartitioned standing variation, with no obvious role for newly formed mutations. © 2023, The Author(s).