Abstract: DNA metabarcoding, as an efficient, cost-effective, and high-resolution tool, has been widely applied in monitoring marine zooplankton diversity and investigating its spatiotemporal distribution patterns. However, compared to its application in coastal environments, the suitability of this technique in the taxonomically complex communities of the tropical open oceans has not been fully evaluated. In this study, vertically stratified zooplankton samples were collected from the tropical eastern Indian Ocean, specifically from the 90°E Ridge and its adjacent waters, using a medium-sized MultiNet plankton sampler targeting two depth layers: 200-500 m and 500-2 000 m. Zooplankton species composition and abundance were analyzed and compared using both mitochondrial COI gene-based DNA metabarcoding (COI metabarcoding) and traditional morphological identification methods, aiming to assess the effectiveness of DNA metabarcoding in zooplankton biodiversity research in the deep ocean. The results showed that COI metabarcoding identified a total of 76 zooplankton species, while morphological identification revealed 92 species. The two approaches showed relatively high consistency in zooplankton species richness across sampling stations, together with substantial overlap at the genus level within the dominant order Calanoida, indicating that COI metabarcoding is effective for species identification of deep-sea zooplankton. However, no significant correlation was found between the relative proportion of sequencing read obtained via metabarcoding for various zooplankton groups and their relative abundance determined by morphological counts. Furthermore, both methods consistently revealed significant vertical differences in zooplankton community structure in the tropical Eastern Indian Ocean.