Taxonomic and nomenclatorial revision of the Neotropical tiger beetle genus Phaeoxantha Chaudoir, 1850 (originally Megacephala Latreille, 1802) is presented. Ammosia Westwood, 1852 (with type species Megacephala bifasciata) is confirmed as a junior synonym of Phaeoxantha. The results by Naviaux (2008), who rectified commonly confused taxonomy and nomenclature of Phaeoxantha nocturna (Dejean, 1831) and P. laminata (Perty, 1830) versus P. limata, based on Megacephala limata Perty, 1833 which is an unjustified emendation by Perty (1833) in Perty (18301833), are confirmed the date of the unjustified emendation is newly rectified here. Megacephala laminata Perty, 1830 is confirmed as the type species of Phaeoxantha. Lectotype of Megacephala nocturna Dejean, 1831 is designated here, based on syntypes from the Dejean-Chaudoir collection in MNHN. The genus is subdivided here into two clearly differentiated subgenera: Phaeoxantha (Phaeoxantha) and Phaeoxantha (Euphaeoxantha) subgen. nov. The nominotypical subgenus is represented by Phaeoxantha laminata (Perty, 1830), P. (P.) nocturna (Dejean, 1831), P. (P.) nocturna crassipunctata ssp. nov., P. (P.) paranocturna sp. nov., P. (P.) epipleuralis W. Horn, 1923, P. (P.) tremolerasi (W. Horn, 1909), P. (P.) cruciata (Brull, 1837) and P. (P.) bifasciata (Brull, 1837). The latter, based on Megacephala bifasciata Brull, 1837, was published by Brull in order to rectify the concept of Megacephala aequinoctialis sensu auctorum, (primarily sensu Dejean 1825, 1833, 1836), which has been commonly yet incorrectly treated in literature as Phaeoxantha aequinoctialis aequinoctialis (Dejean, 1825). Although Dejean (1825) clearly characterized the same Megacephala species, his act cannot be interpreted as a valid description under Article 12 of the ICZN (1999), because he explicitly referred to the publications by Linnaeus (1763) and redescription by Fabricius (1775). Therefore, his act must be interpreted as a misidentification and subsequent usage of the name Cicindela aequinoctialis Linnaeus, 1763, which is in fact a bombardier beetle (Brachinini, Carabidae), presently known as Pheropsophus aequinoctialis (Linnaeus, 1763), misidentified by Dejean as Megacephala. In contrast, the name Megacephala bifasciata Brull, 1837 was validly published, supported by a rather appropriate illustration and preserved type specimen, being also in common use to the present day; therefore, the confused name aequinoctialis cannot be preserved and is considered an unavailable name being excluded here from Megacephalini and Cicindelidae. The new subgenus Phaeoxantha (Euphaeoxantha) subgen nov. (type species Megacephala testudinea Klug, 1834) comprises following five species: P. (E.) testudinea (Klug, 1834), P. (E.) klugii (Chaudoir, 1850), P. (E.) bucephala (W. Horn, 1909), P. (E.) wimmeri (Mandl, 1958) and P. (E.) lindemannae (Mandl, 1964). Keys to the two subgenera and to their species, descriptions or differential diagnoses and colour photographs of the habitus and of characters of available type specimens of all species of the genus, as well as the variability and characters of individual populations, are presented and their distribution discussed.