Examination of fungal specimens collected in the Atlantic rain forest ecosystems of Northeast Brazil revealed many potentially new epigeous and semihypogeous glomerocarp-producing species of the phylum Glomeromycota. Among them were two fungi that formed unorganized epigeous glomerocarps with glomoid spores of almost identical morphology. The sole structure that distinguished the two fungi was the laminate layer 2 of their three-layered spore wall, which in spores of the second fungus crushed in PVLG-based mountants contracted and, consequently, transferred into a crown-like structure. Surprisingly, phylogenetic analyses of sequences of the 18S-ITS-28S nuc rDNA and the rpb1 gene indicated that these glomerocarps represent two strongly divergent undescribed species in the family Glomeraceae. The analyses placed the first in the genus Dominikia, and the second in a sister clade to the monospecific generic clade Kamienskia with Kamienskia bistrata. The first species was described here as Dominikia glomerocarpica sp. nov. Because D. glomerocarpica is the first glomerocarp-forming species in Dominikia, the generic description of this genus was emended. The very large phylogenetic distance and the fundamental morphological differences between the second species and K. bistrata suggested us to introduce a new genus, here named as Epigeocarpum gen. nov., and name the new species Epigeocarpum crypticum sp. nov. In addition, our analyses also focused on an arbuscular mycorrhizal fungus originally described as Rhizophagus neocaledonicus, later transferred to the genus Rhizoglomus. The analyses indicated that this species does not belong to any of these two genera but represents a new clade at the rank of genus in the Glomeraceae, here described as Silvaspora gen. nov.