Reverse breeding is a new plant breeding strategy based on crossover suppression during meiosis. If crossover formation is suppressed in a heterozygous plant, this brings forth unprecedented possibilities like the almost instantaneous generation of homozygous parents for that heterozygote. As a proof of concept, we generated an Arabidopsis (Columbia-Landsberg) heterozygote that carried a RNAi:DMC1 construct that knocks-down crossover formation. Gametes of this heterozygote were grown into doubled haploid offspring. These offspring consisted of a variety of plants showing different combinations of (non-recombinant) Columbia and Landsberg chromosomes. Among these we retrieved the original Columbia parent and a complete set of chromosome substitution lines. Because the number of possible chromosome combinations is finite, we could easily select two so called "complementing DHs" from which the Col-Ler hybrid could be re-created. Essentially, breeders can now bring single choice uncharacterized heterozygotes (like F2's or plants from outcrossing populations) into a hybrid breeding program by creating parental lines for them. Reverse breeding superficially resembles apomixis (clonal reproduction through seeds) since both allow the preservation of heterozygous genotypes. Reverse breeding, however, has very different uses because it generates homozygous breeding lines. It thus allows for the improvement of the starting heterozygote because new traits can be introgressed into its newly produced parental lines. Reverse breeding is thought to be suitable for crops with smaller chromosome numbers (x ≤ 12). It will be discussed how reverse breeding could be develped for such crops, and it will be shown how reverse breeding presents very interesting new possibilities studying epistasis and heterosis through chromosome substitution lines.