Lycidae, net winged beetles, have proved difficult to classify using morphological characters. Here, using a previously published molecular phylogeny, comparing the results with morphological data and re-analyzing previously published morphological data set, we propose a revised classification of Lycidae. All analyses support the monophyly of Lycidae, but phylogeny inferred from molecular data is in conflict with the current classification. The adult larviform females evolved in several lineages and never switched back to a winged form. Therefore, neotenic development of females is not a synapomorphy of Lycidae and the neotenic lineages do not form a basal paraphylum with respect to remaining Lycidae as previously proposed. As a consequence morphological similarities resulting from neoteny are homoplasies and cannot be used for definition of monophyletic lineages. The major result of this study is delineation of five basal clades, which are given subfamily rank: Libnetinae Bocak et Bocakova, 1990, stat. nov., Dictyopterinae Kleine, 1928, stat. nov., Lyropaeinae Bocak et Bocakova, 1989, Ateliinae Kleine, 1928 and Lycinae Laporte, 1836, sensu nov. Dexorinae Bocak et Bocakova, 1989, stat. nov. were not available for molecular analyses and their position is inferred from morphology alone. Further, Lycinae are redefined. Leptolycini are nested among Neotropical Lycini and render them paraphyletic. Slipinskiini are shown to group with Erotini. The distant position of Macrolycini and Dilophotini is demonstrated and Dilophotini are transferred from Calochrominae to Ateliinae. Four new tribes are proposed: Lycoprogenthini trib. nov. in Dictyopterinae, Alyculini trib. nov. and Antennolycini trib. nov. in Lyropaeinae, Dihammatini trib. nov. in Lycinae. Calochromini Lacordaire, 1857, stat. nov. and Leptolycini Leng et Mutchler, 1922, stat. nov. are lowered to the tribal status in Lycinae. The validity of Thonalmini Kleine, 1933 stat. rev. is re-established.