Parasenecio with c. 70 species is a major component of the eastern Asian Tussilagininae (Asteraceae, Senecioneae). Previous molecular phylogenetic studies, although with limited taxon sampling, have suggested that it is polyphyletic and forms a complex with ten other eastern Asian genera of Tussilagininae, i.e. the Ligularia–Cremanthodium–Parasenecio (LCP) complex. Here we sample Parasenecio and its relatives in the LCP complex densely and generate two phylogenetic hypotheses for the genus using four plastid DNA (trnCGCA-petN, trnGUCC-trnTGGU-psbD, trnTUGU-trnLUAA and rpl32-ndhF) and two nuclear ribosomal (nr)DNA [internal and external transcribed spacer (ITS and ETS)] regions, respectively. Analyses of the plastid DNA data produce a robust generic-level phylogenetic tree for Parasenecio and clearly demonstrate the polyphyletic pattern of the genus. Most notably, two sections of Parasenecio, P. section Delphiniifolii s.s. and P. section Taimingasa, are shown to be distantly related to core Parasenecio, being sister to Miricacalia and Syneilesis, respectively, with strong support. The small genus Sinacalia is also revealed to be polyphyletic, with S. caroli nested in core Parasenecio, and S. tangutica and S. davidii forming a clade sister to the core Parasenecio–S. caroli clade. The nrDNA tree is, however, relatively poorly resolved, only confirming the sister relationship between P. section Delphiniifolii s.s. and Miricacalia. Sinacalia s.s. is resolved as being deeply nested in core Parasenecio with nrDNA data, which is the only significant incongruence with the plastid DNA hypothesis relevant to the generic circumscription of Parasenecio. Statistical topology (approximately unbiased and Swofford–Olsen–Waddell–Hillis) tests suggest that, except for the position of Sinacalia s.s., the generic-level relationships of Parasenecio inferred from the plastid DNA data are not incongruent with those from the nrDNA data and, on the basis of current sequence data, the inferred molecular phylogenetic relationships should be preferred over the current classification. Based on these results and evidence from morphology and karyology, we propose a new circumscription of Parasenecio, involving transferring S. caroli to Parasenecio as P. caroli comb. nov. and segregating P. section Delphiniifolii s.s. and P. section Taimingasa as the new genera Japonicalia gen. nov. and Taimingasa gen. nov., respectively. Sinacalia s.s., with significant incongruence between the plastid DNA and nrDNA trees, is tentatively retained as an independent genus pending future studies. A revised key to Parasenecio and its close relatives is provided.