More than 2,500 species of copepods (Class Maxillopoda; Subclass Copepoda) occur in the marine planktonic environment. The exceptional morphological conservation of the group, with numerous sibling species groups, makes the identification of species challenging, even for expert taxonomists. Molecular approaches to species identification have allowed rapid detection, discrimination, and identification of species based on DNA sequencing of single specimens and environmental samples. Despite the recent development of diverse genetic and genomic markers, the barcode region of the mitochondrial cytochrome c oxidase subunit I (COI) gene remains a useful and - in some cases - unequaled diagnostic character for species-level identification of copepods. This study reports 800 new barcode sequences for 63 copepod species not included in any previous study and examines the reliability and resolution of diverse statistical approaches to species identification based upon a dataset of 1,381 barcode sequences for 195 copepod species. We explore the impact of missing data (i.e., species not represented in the barcode database) on the accuracy and reliability of species identifications. Among the tested approaches, the best close match analysis resulted in accurate identification of all individuals to species, with no errors (false positives), and out-performed automated tree-based or BLAST based analyses. This comparative analysis yields new understanding of the strengths and weaknesses of DNA barcoding and confirms the value of DNA barcodes for species identification of copepods, including both individual specimens and bulk samples. Continued integrative morphological-molecular taxonomic analysis is needed to produce a taxonomically-comprehensive database of barcode sequences for all species of marine copepods.