In sexual organisms, recombination suppression can evolve in specific genomic regions to protect beneficial allelic combinations, resulting in the transmission of multiple genes as a single locus, which is called a supergene. Supergenes determine complex phenotypes, such as gender in organisms with sex chromosomes. Some sex chromosomes display successive steps of recombination suppression known as “evolutionary strata”, which are commonly thought to result from the successive linkage of sexually antagonistic genes (i.e. alleles beneficial to one sex but detrimental to the other) to the sex-determining region. There has however been little empirical evidence supporting this hypothesis. Fungi constitute interesting models for studying the evolutionary causes of recombination suppression in sex-related chromosomes, as they can display non-recombining mating-type chromosomes not associated with male/female functions. Here, we studied the evolution of recombination suppression on mating-type chromosomes in the Microbotryum plant-castrating fungi using comparative genomic approaches. In Microbotryum fungi, mating occurs between gametes with distinct alleles at the two mating-type loci, as is typical of basidiomycete fungi. We showed that recombination suppression evolved multiple times independently to link the two mating-type loci from an ancestral state with mating-type loci on two distinct chromosomes. Recombination suppression either linked the mating-type genes to their respective centromere or linked mating-type loci after they were brought onto the same chromosome through genomic rearrangements that differed between species. Both types of linkage are beneficial under the intra-tetrad mating system of Microbotryum fungi as they increase the odds of gamete compatibility. Recombination suppression thus evolved multiple times through distinct evolutionary pathways and distinct genomic changes, which give insights about the repeatability and predictability of evolution. We also reported the existence of independent evolutionary strata on the mating-type chromosomes of several Microbotryum species, which questions the role of sexual antagonism in the stepwise extension of non-recombining regions because mating-types are not associated with male/female functions. Previous studies reported little phenotypic differences associated to mating-types, rending unlikely any antagonistic selection between mating types (i.e. “mating-type antagonism”, with genes having alleles beneficial to one mating-type but detrimental to the other). The genes located in non-recombining regions on the mating-type chromosomes can be differentially expressed between mating types, but our analyses indicated that such differential expression was more likely to result from genomic degeneration than from mating-type antagonism. Deleterious mutations are indeed known to accumulate in non-recombining regions resulting in modifications of gene expression or of protein sequence. We concluded that antagonistic selection cannot explain the formation of evolutionary strata in Microbotryum fungi. Alternative mechanisms must be therefore be considered to explain the stepwise expansion of non-recombining regions, and they could also be important on sex chromosomes. This work thus prompts for future studies to identify further evolutionary strata not associated with male/female functions as well as to elucidate their evolutionary causes and consequences in terms of genomic degeneration.