The phylogenetic relationships within and among subtribes of the fruit fly tribe Dacini (Ceratitidina, Dacina, Gastrozonina) were investigated by sequencing four mitochondrial and one nuclear gene fragment. Bayesian, maximum likelihood and maximum parsimony analyses were implemented on two datasets. The first, aiming at obtaining the strongest phylogenetic signal (yet, having lower taxon coverage), consisted of 98 vouchers and 2338 concatenated base pairs (bp). The second, aiming at obtaining the largest taxonomic coverage (yet, providing lower resolution), included 159 vouchers and 1200 concatenated bp. Phylogenetic relationships inferred by different tree reconstruction methods were largely congruent and showed a general agreement between concatenated tree topologies. Yet, local conflicts in phylogenetic signals evidenced a number of critical sectors in the phylogeny of Dacini fruit flies. All three Dacini subtribes were recovered as monophyletic. Yet, within the subtribe Ceratitidina only Perilampsis and Capparimyia formed well-resolved monophyletic groups while Ceratitis, and Trirhithrum did not. Carpophthoromyia was paraphyletic because it included Trirhithrum demeyeri and Ceratitis connexa. Complex phylogenetic relationships and localised conflict in phylogenetic signals were observed within subtribe Dacina with (a) Dacus, (b) the Bactrocera subgenus Zeugodacus and (c) all other Bactrocera species forming separate clades. The subgenus Bactrocera (Zeugodacus) is therefore raised to generic rank (Zeugodacus Hendel stat. nov.). Additionally, Bactrocera subgenera grouped under the Zeugodacus group should be considered under new generic combinations. Although there are indications that Zeugodacus and Dacus are sister groups, the exact relationship between Zeugodacus stat. nov., Dacus and Bactrocera still needs to be properly resolved.