The evolution of anopheline mosquitoes (Culicidae: Anophelinae) has been the subject of speculation and study for decades, but a comprehensive phylogeny of these insects is far from complete. The results of phylogenetic studies based on morphological and molecular data sets are conspicuously ambiguous. Here, we revisit the phylogenetic relationships of anopheline mosquitoes using state-of-the-art software and cladistic methods to analyse the data set of Harbach & Kitching (2005). We present a refined interpretation of relationships based on analyses of a revised data set that includes an additional species. Implied weighting analyses were conducted with TNT with the concavity constant K ranging from 1 to 33. We determined the optimal K value by summing the GC supports for each MPC and selected the tree with the highest support, K = 30, as the preferred cladogram. We then collapsed the branches with GC support < 1 to obtain the 'best' topography of relationships. Genus Chagasia is the basalmost taxon of Anophelinae, and genus Anopheles is recovered as monophyletic but only if Anopheles implexus is excluded and genus Bironella is subordinated within it. The Afrotropical An. implexus is recovered as the sister to all other anophelines, and Christya Theobald, stat. nov., is elevated from synonymy with Anopheles Meigen as a subgenus to accommodate it. The other anophelines comprise two large clades. The first includes the reciprocally monophyletic subgenera Kerteszia + Nyssorhynchus; the second consists of subgenus Cellia as the sister to a heterogeneous clade that includes genus Bironella and subgenera Anopheles, Baimaia, Lophopodomyia and Stethomyia of genus Anopheles. The sister relationship of Cellia and the heterogeneous clade is lost when the branches with GC < 1 are collapsed. The monophyly and non-monophyly of the informal subordinate taxa of subgenera Nyssorhynchus, Cellia and Anopheles, and also evolutionary scenarios, are discussed in relation to previous studies.