Typhuloid fungi are a very poorly known group of tiny clavarioid homobasidiomycetes. The phylogenetic position and family classification of the genera targeted here, Ceratellopsis, Macrotyphula, Pterula sensu lato and Typhula, are controversial and based on unresolved phylogenies. Our six-gene phylogeny with an expanded taxon sampling shows that typhuloid fungi evolved at least twice in the Agaricales (Pleurotineae, Clavariineae) and once in the Hymenochaetales. Macrotyphula, Pterulicium and Typhula are nested within the Pleurotineae. The type of Typhula (1818) and Sclerotium (1790), T. phacorrhiza and S. complanatum (synonym T. phacorrhiza), are encompassed in the Macrotyphula clade that is distantly related to a monophyletic group formed by species usually assigned to Typhula. Thus, the correct name for Macrotyphula (1972) and Typhula is Sclerotium and all Typhula species but those in the T. phacorrhiza group need to be transferred to Pistillaria (1821). To avoid undesirable nomenclatural changes, we suggest to conserve Typhula with T. incarnata as type. Clavariaceae is supported as a separate, early diverging lineage within Agaricales, with Hygrophoraceae as a successive sister taxon to the rest of the Agaricales. Ceratellopsis s. auct. is polyphyletic because C. acuminata nests in Clavariaceae and C. sagittiformis in the Hymenochaetales. Ceratellopsis is found to be an earlier name for Pterulicium, because the type, C. queletii, represents Pterulicium gracile (synonym Pterula gracilis), deeply nested in the Pterulicium clade. To avoid re-combining a large number of names in Ceratellopsis we suggest to conserve it with C. acuminata as type. The new genus Bryopistillaria is created to include C. sagittiformis. The families Sarcomyxaceae and Phyllotopsidaceae, and the suborder Clavariineae, are described as new. Six new combinations are proposed and 15 names typified.