• Assortative mating;
  • asymmetrical sexual isolation;
  • meta-analysis;
  • patterns of speciation in Drosophila;
  • reproductive character displacement;
  • secondary contact;
  • Templeton effect

Understanding how often natural selection directly favors speciation, a process known as reinforcement, has remained an outstanding problem for over 70 years. Although reinforcement has been strongly criticized in the past, it is once again seen as more realistic due to the seminal discovery of enhanced prezygotic isolation among sympatric species and to a handful of well-studied examples. Nevertheless, the pattern of enhanced isolation in sympatry has alternative explanations, highlighting the need to uncover unique signatures of reinforcement to determine its overall frequency in nature. Using a novel dataset on asymmetrical prezygotic and postzygotic isolation among Drosophila species, I uncover new patterns explicitly predicted by reinforcement. Broadly, I found that almost all sympatric species had concordant isolation asymmetries, where the more costly reciprocal mating has greater prezygotic isolation relative to the less costly mating. No such patterns exist in allopatry. Using simulations, I ruled out alternative explanations and showed that concordant isolation asymmetries in sympatry are likely unique signatures of reinforcement. These results allowed me to estimate that reinforcement may impact 60–83% of all sympatric Drosophila and enhance premating isolation by 18–26%. These findings suggest that reinforcement plays a key role in Drosophila speciation.