Despite the growing use and potential of DNA metabarcoding to improve and expedite macrozoobenthos monitoring, its employment in Water Framework Directive (WFD) monitoring of transitional ecosystems still remains largely unexplored and pending proof-of-concept studies. In the current study, we addressed this research gap by building upon the biomonitoring network program of the Portuguese Environmental Agency (APA) to benchmark metabarcoding against the morphology-based approach for characterizing macrozoobenthic communities. We assessed the ecological condition of 20 sites from four major transitional ecosystems in the west coast of Portugal, namely Minho, Lima, Vouga and Mondego estuaries. A total of 154 marine invertebrate species were detected with both methodologies, distributed by 11 phyla. In the majority of the sites, metabarcoding returned a higher number of species and phyla than the morphology-based approach (up to 2.5 times higher). In parallel, the proportion of species detected concurrently by both methods was low (35 species, 23%). The use of a multi-locus strategy increased recovered diversity through metabarcoding, since 37 species were detected exclusively with COI and 46 with 18S. For about 61% of the species recovered through morphology, metabarcoding failed detection, among which 20% was due to the lack of reference sequences in genetic databases. For the remaining, we did not find any plausible reason for only 10%, which could be due either to inefficient DNA extraction or PCR failure. Although morphological and metabarcoding-derived biotic indices did not match completely, similar responses to the environmental gradient were obtained in morphology and metabarcoding based-datasets. We anticipate that metabarcoding can increase the throughput and quality of the assessments, allowing faster assessments with greater spatial-temporal density and robust identifications of all specimens in a sample including larval stages, juveniles, and cryptic lineages, as well as smaller taxonomic groups that cannot be identified to species level using the traditional approach.