The phylogeny of the basal hymenopteran lineages, including representatives of all ‘symphytan' families, is analysed. In total, 236 morphological characters were scored for 44 exemplars, including six outgroup, two xyelid, 16 tenthredinoid, five pamphilioid, three cephoid, five ‘siricoid', one orussid, and six apocritan taxa. The dataset is analysed with parsimony under equal weights and under implied weights. The monophyly of the Hymenoptera is strongly supported but the sistergroup of the Hymenoptera cannot be identified with confidence. The relationships of the ‘symphytan' lineages are Xyeloidea+(Tenthredinoidea+(Pamphilioidea+(Cephoidea+(Anaxyelidae+ (Siricidae+(Xiphydriidae+(Orussoidea+Apocrita))))))). Many of the relationships between the superfamilies, especially in the basal branching pattern, are only weakly corroborated. The monophyly of most superfamilies is well supported, and all may be monophyletic except the ‘Siricoidea', which is clearly paraphyletic. It is difficult to decide whether the Siricidae or the Anaxyelidae are the closest relatives of Xiphydriidae+(Orussoidea+Apocrita). The support for the sistergroup relationship between the Orussoidea and the Apocrita is substantial, putative synapomorphies being provided by most character systems. There is also good evidence in favour of the monophyly of the Apocrita. The internal phylogeny of the Tenthredinoidea differs considerably from the results of earlier analyses. The Blasticotomidae are the sistergroup of the Tenthredinoidea s.s. Relationships at the base of the Tenthredinoidea are weakly supported. It is uncertain whether the Tenthredinidae are monophyletic or comprise a basal paraphyletic grade within the Tenthredinoidea s.s. The Diprionidae may be the sistergroup to the Cimbicidae+(Argidae+Pergidae). Most relationships within the Cimbicidae+(Argidae+Pergidae) clade are well  corroborated, with the exception of the monophyly of the Argidae. It is proposed to elevate the Anaxyelidae and the Xiphydriida both to superfamily status. The family-level classification of the Tenthredinoidea will probably also have to be changed, but this must await further clarification of the phylogeny of this superfamily.