With around 2,000 species, Carex (Cyperaceae) is the second-largest genus of monocots after Bulbophyllum (Orchidaceae). The traditional classification system of Carex, based on morphology, has recently been updated by the Global Carex Group (GCG). It was based on two detailed molecular studies: 1) an analysis that sampled many species (996 species) using a few genetic markers (matK, ITS, ETS-1f), and 2) a study that sampled fewer species (88 species) but used a considerable amount of data (308 nuclear exon, 543 intron, and 66 plastid exon regions) obtained via a Hyb-Seq method. However, this new classification system has not yet been evaluated with respect to an extensive plastid genomic dataset. Our previous research determined that complete plastid genomes in Carex could not be aligned due to significant structural differences among species. Therefore, we focused our analyses on 82 coding regions in a taxonomically representative sample (187 species) of the subgroups defined in the new classification system for Carex. Results showed that four out of the six subgenera proposed in the new classification system were highly supported (BS=100%) in the maximum-likelihood tree. Moreover, of 62 formally named Linnean sections and 49 informal groups that were sampled for more than one species, 70 % were highly supported (BS>80%). However, the relationships of subgenus Vignea differed in plastid and nuclear trees, suggesting a possible hybrid origin for this subgenus. Additionally, we observed many conflicting relationships among sections and informally recognized groups suggesting that these lineages may also have ancient hybrid origins. Our study reveals that plastid data largely supports the new classification system proposed for Carex but highlights conflicting relationships among some subgroups that could be due to the different evolutionary histories of plastid and nuclear genomes.