Fish eggs, larvae, and juveniles are important food sources for fish populations. Correct identification of fish eggs, larvae, and juveniles can let us understand which species are spawning where and when, their hatching and nursery grounds, and their possible migration routes during their early life history. At present, fish larval and juvenile identification mostly relies on morphological characters. But since different species often share similar characters and species change greatly during their early development stages, identification of fish larvae and juveniles is very hard, most of which can only be identified to the family or genus level, except for a few fish juveniles that can be identified to species. Moreover, because traditional morphological methods rely too much on personal experience and observation methods, it is easy to misidentify fish larvae and juveniles and difficult to identify to the genus or species level. In view of the limitations of traditional morphological methods, DNA barcode techniques could be a rapid tool to survey many uncertain species, species composition, and cryptic species and to distinguish morphologically similar species. To study the applicability of this technique to identify fish larvae and juveniles, samples were collected in Fujian coastal areas from July 17 to 20, 2015, and 80 samples were selected for DNA barcoding analysis. A total of 73 valid cytochrome C oxidase subunit I (CO I) gene sequences of fish larvae and juveniles in Fujian coastal areas were obtained. Five species were identified to the level of genus, two were identified to the level of family, and another 26 species were identified to the level of species. The results showed that the average Kimura-2-parameter (K2P) distances within species, genus, family, and order were 0.0023, 0.1797, 0.1937, and 0.2420, respectively. The interspecific genetic distance was much larger than the intraspecific distance, and K2P genetic distance values increased with taxonomic level. All sequences formed species units in a neighbor-joining phylogenetic tree, indicating that the DNA barcode can be used to identify these 33 species. Consequently, CO I barcoding is one of the useful tools to identify fish larvae and juveniles. However, this test highlights the limitations of DNA barcoding. The neighbor-joining tree cannot analyze cluster relations clearly above the taxonomic level of family, which corresponds to the slower-increasing trend of genetic distance values with the higher taxa. This means that CO I barcoding is only suitable for phylogenetic analysis of lower taxa, and is not applicable for species identification of higher taxa. Four sequences having high similarity (greater than 98%) with many species in the database were identified only to genus ( sp.). That indicates that some fish species are similar not only in shape but also in CO I sequences, and sequence identification cannot be done using only CO I barcoding. We should strengthen the research of multi-gene barcode technology in order to overcome the deficiency of the single CO I gene sequence. Another sequence, limited by the number of sequences that can be used in the DNA database, with a maximum similarity between 85%-86%, was identified to the level of family. It is suggested to further improve the sequence quantity and quality of the DNA barcode database, so as to let it be more useful for species identification.