Zooplankton are key groups in marine ecosystems, which link primary producers and higher tropic levels. Their assemblages are substantially affected by environmental changes. The monitoring of large-scale and rapid changes in zooplankton biodiversity is of great importance to provide guidelines for marine fishery management and ecosystem protection. Zooplankton are widely distributed in the global ocean, with systematic complexity and high biodiversity. Therefore, the traditional morphology-based identification methods largely depend on sufficient taxonomic skills of experts, which costs more time, money, and expertise. Moreover, morphological identification provides a significantly lower estimation of zooplankton biodiversity due to the numerous cryptic species. These factors limit our understanding of biodiversity and ecology of zooplankton. DNA metabarcoding is proved to be a promising and powerful tool for species and community monitoring. Instead of separating individual organisms, this new approach extracts total DNA from the mixed samples of zooplankton and rely on high-throughput sequencing platform to enable rapid, accurate, and cost-effective analysis of a large number samples. The considerable information generated in the high-throughput sequencing process is helpful to reveal the hidden diversity of marine zooplankton community. In addition to samples collected by plankton nets, the environmental DNA (eDNA) from water column could also be used for metabarcoding analysis. The use of eDNA combined with metabarcoding method to detect biodiversity is less destructive to the environment and zooplankton community, and could be a complementary approach to detect species that are hardly be captured or easily destroyed by trawls. Thus, this approach is increasingly being used. In this review, the ribosome and mitochondrial molecular markers, which were frequently used in recently published studies related to DNA metabarcoding of zooplankton, were listed and compared. Particularly, small subunit ribosomal RNA (18S) and mitochondrial cytochrome oxidase I (COI) genes are mostly used. The former has higher taxonomic coverage, while the latter has higher taxonomic resolution. Since the coverage of individual markers is not sufficient by lack of reference sequences in the database, the use of multiple barcodes combined are recommend. In addition, the reliability and shortcomings of these technologies used in the study of zooplankton ecology were discussed, comparing with the results of morphological technology. Then some aspects of application, such as the monitoring of marine zooplankton community, trophic relationship analysis, and early warning of biological invasions, were reviewed. Finally, we prospected that the mutli-gene barcodes, complete reference libraries and accurate quantification research are important directions for the development of DNA metabarcoding approaches in the future.