Abstract The Old World sparrows include some of the best-studied passerine species, such as the cosmopolitan human commensal, the house sparrow (Passer domesticus) as well as poorly studied narrow-range endemics like the Iago sparrow (P. iagoensis) from the Cape Verde Archipelago or specialists from extreme environments like the desert sparrow (P. simplex). It is therefore notable that to date the most complete phylogenetic hypothesis for the Old World sparrows comprised only ten of 43 currently accepted species. With this study we provide an updated phylogeny of Passeridae covering about two third of the family’s species richness. Though still being far from taxon-complete, this new phylogenetic hypothesis provides firm evidence to clarify some open taxonomic questions. All genus-level taxa were reciprocally monophyletic with strong support. Contrary to previous classifications, bush sparrows and rock sparrows were not sister taxa, and therefore their classification in separate genera Gymnoris and Petronia is justified. Plumage color traits like the yellow throat patch of the latter two genera or head color pattern in Passer species do not provide reliable phylogenetic information, except for the large-sized African grey-headed sparrows that resulted as a monophyletic group (P. diffusus, P. griseus, P. gongoensis). Unexpectedly, two small-sized species, P. eminibey and P. luteus that to date are regarded as close relatives were firmly nested in two separate clades of Passer sparrows. Therefore, their separate generic treatment under Sorella eminibey and Auripasser luteus (together with A. euchlorus) does not seem justified.