BACKGROUND: Potentilla L. and Dasiphora L. are predominantly perennial herbs, occasionally manifesting as annuals or shrubs, primarily found in the northern temperate zone. However, taxonomic classification within this group remains contentious, particularly regarding genus boundaries and species delineation. Therefore, this study sequenced and analyzed the complete plastid genomes of 19 species from Potentilla and Dasiphora, comparing them with five previously published plastid sequences. Our objectives included reconstructing phylogenetic relationships within Potentilla and Dasiphora and investigating cytonuclear discordance among them. RESULTS: These plastid genomes were highly conserved in structure, GC content, and overall genome composition, comprising 84 protein-coding genes, 37 tRNA genes, and 8 rRNA genes. Notably, all Dasiphora plastid genomes lacked the unique intron for rpl2. Comparative genomic analyses revealed that variations in plastid genome size were due to differences in the lengths of the LSC, SSC, and IR regions. The IR region was predominantly conserved, while non-coding regions exhibited higher variability than coding regions. We screened SSR and identified seven highly variable loci that serve as potential molecular markers, offering valuable insights into the intergeneric relationships between Potentilla and Dasiphora. Phylogenetic analyses based on nuclear (ITS, ETS) and cytoplasmic (plastid, mitochondrial) genes confirmed the monophyly of Potentilla and Dasiphora, with results largely consistent with previous studies and supported by robust reliability metrics. We identified cytonuclear conflicts within Potentilla, which frequently disrupt its monophyly. We hypothesize that these conflicts may result from interspecific hybridization or incomplete lineage sorting events during the evolutionary history of the genus. CONCLUSIONS: This study offers a theoretical foundation for advancing molecular identification and phylogenetic research on Potentilla and Dasiphora species. However, future work could benefit from greater detail on the criteria for selecting mitochondrial gene sequences and nrDNA datasets.