We present a molecular phylogeny of Nitidulidae based on thirty ingroup taxa representing eight of the ten currently recognized subfamilies. Approximately 10 K base pairs from seven loci (12S, 16S, 18S, 28S, COI, COII and H3) were used for the phylogenetic reconstruction. The phylogeny supports the following main conclusions: (i) Cybocephalidae are formally recognized as a distinct family not closely related to Nitidulidae and its constituent taxa are defined; (ii) Kateretidae are sister to Nitidulidae; (iii) Cryptarchinae are monophyletic and sister to the remaining nitidulid subfamilies; (iv) subfamily Prometopinae stat. res. is reinstated and defined, to accommodate taxa allied to Axyra Erichson, Prometopia Erichson and Megauchenia MacLeay; (v) Amphicrossinae, Carpophilinae and Epuraeinae are shown to be closely related taxa within a well-supported monophyletic clade; (vi) tribal affinities and respective monophyly within Nitidulinae are poorly resolved by our data and must be more rigorously tested as there was little or no support for prior morphologically based tribes or genus-level complexes; (vii) Nitidulinae are found to be paraphyletic with respect to Cillaeinae and Meligethinae, suggesting that they should either be subsumed as tribes, or Nitidulinae should be divided into several subfamilies to preserve the status of Cillaeinae and Meligethinae; (viii) Teichostethus Sharp stat. res. is not a synonym of Hebascus Erichson and the former is reinstated as a valid genus. These conclusions and emendations are discussed in detail and presented within a morphological framework.