The soft scales (Hemiptera: Coccoidea: Coccidae) are a group of sap-sucking plant parasites, many of which are notorious agricultural pests. The quarantine and economic importance of soft scales necessitates rapid and reliable identification of these taxa. Nucleotide sequences of the mitochondrial cytochrome c oxidase subunit I (COI) gene (barcoding region) and 28S rDNA were generated from 340 individuals of 36 common soft scales in China. Distance-based [(best match, Automated Barcode Gap Discovery (ABGD)], tree-based (neighbor-joining, Bayesian inference), Klee diagrams, and general mixed Yule coalescent (GMYC) models were used to evaluate barcoding success rates in the data set. Best match showed that COI and 28S sequences could provide 100 and 95.52% correct identification, respectively. The average interspecific divergences were 19.81% for COI data and 20.38% for 28S data, and mean intraspecific divergences were 0.56 and 0.07%, respectively. For COI data, multiple methods (ABGD, Klee, and tree-based methods) resulted in general congruence with morphological identifications. However, GMYC analysis tended to provide more molecular operational taxonomic units (MOTUs). Twelve MOTUs derived from five morphospecies (Rhodococcus sariuoni, Pulvinaria vitis, Pulvinaria aurantii, Parasaissetia nigra, and Ceroplastes rubens) were observed using the GMYC approach. In addition, tree-based methods showed that 28S sequences could be used for species-level identification (except for Ceroplastes ceriferus – Ceroplastes pseudoceriferus), even with low genetic variation (<1%). This report demonstrates the robustness of DNA barcoding for species discrimination of soft scales with two molecular markers (COI and 28S) and provides a reliable barcode library and rapid diagnostic tool for common soft scales in China.