The genus Phyllogomphus Selys, 1854 inhabits tropical Africa, largely allopatric to Ceratogomphus Selys, 1854 of southern Africa and Isomma Selys, 1892 (including its probable synonym Malgassogom- phus Cammaerts, 1987) of Madagascar. Together these genera form the distinctive Afrotropical sub- family Phyllogomphinae (Carle 1986). Despite the revision of Phyllogomphus by Fraser (1957), the genus is in urgent need of taxonomic attention (Dijkstra 2003, Vick 1999). Until now 22 names have been introduced, but they probably represent less than half that number of species (table 1). Phyllogomphus can be separated into two groups: a) species with five stripes on each side of the synthorax and b) species with the antehumeral and metepisternal stripes ab- sent or strongly reduced, thus leaving only three tho- racic stripes. The five-striped species occur from The Gambia to Nigeria. The three-striped species occur further east, ranging from Bioko and Cameroon to Kenya and South Africa. The present paper revises the latter group. Much of the confusion within this group originated from Fraser’s (1949, 1957) misin- terpretation of P. selysi. The extensive material labelled as P. selysi in the Musée Royal de l’Afrique Centrale, Tervuren (mrac) (27 males, 9 females) contains two species. Only a small proportion (3 males, 3 fe- males, mostly type material) agrees with Schouteden’s (1933) species. The majority belongs to a different species, which Fraser (1949, 1957) however described as P. selysi. Recently, the first author found P. selysi Schouteden and P. selysi sensu Fraser nec Schouteden occurring together in W Uganda (figs. 27, 29-30). The quest for the correct name of the second species provided the stimulus for this paper. Due to Fraser’s mistake, an over-emphasis on variation and the de- scription of taxa based on females only, sixteen names have been introduced in this group, a quarter with female holotypes. We believe that only four three- striped species exist, which are somewhat variable and widely overlapping in range, size and colour, but mor- phologically very distinctive. 1 Klaas-Douwe B. DIJKSTRA 1 , Viola CLAUSNITZER 2 & Graham S. VICK 3 1 National Museum of Natural History Naturalis, Leiden, The Netherlands 2 Halle/Saale, Germany; 3 Tadley, United Kingdom REVISION OF THE THREE-STRIPED SPECIES OF PHYLLOGOMPHUS (ODONATA, GOMPHIDAE) Dijkstra K.-D.B., V. Clausnitzer & G.S. Vick, 2006. Revision of the three-striped species of Phyllogomphus (Odonata, Gomphidae). – Tijdschrift voor Entomologie 149: 1-14, figs.1-32, tables 1-2. [issn 0040-7496]. Published 1 June 2006. The taxonomy of the Phyllogomphus species occurring from Cameroon eastwards, characterised by three-striped sides of the thorax, has been confused by misinterpretation of the identity of the most widespread species, P. selysi, and substantial variation in the species. Of sixteen named taxa, only four are considered valid species after clarifying the identity of P. selysi, matching females to the correct males, and accounting for variation, particularly of size, colour and the morphology of the vulvar scale: P. annulus is not a synonym of the true P. selysi but of Fraser’s in- terpretation of the latter species; P. dundomajoricus and P. dundominusculus are junior synonyms of P. annulus; P. montanus, P. hartwigi, P. perisi and P. margaritae of P. coloratus; P. orientalis, P. edentatus, P. latifasciae, P. symoensi, P. brunneus and P. corbetae of P. selysi. Keys to the species and distribution maps are provided, and the taxonomy of the genus is discussed. Correspondence: K.-D.B. Dijkstra, National Museum of Natural History Naturalis, PO Box 9517, NL-2300 RA Leiden, The Netherlands. E-mail: dijkstra@nnm.nl E-mail: dijkstra@nnm.nl V. Clausnitzer, Graefestrasse 17, D-06110 Halle/Saale, Germany. E-mail: violacl@t-online.de E-mail: violacl@t-online.de G.S. Vick, Crossfields, Little London, Tadley, Hants RG26 5ET, United Kingdom. E-mail: camdragonfly@aol.com Key words. – Odonata; Dragonfly; Anisoptera; Phyllogomphus; Taxonomy; Synonym; Africa.