High-throughput sequencing (HTS) technologies offer new promise to support surveillance programs targeting marine non-indigenous species (NIS). Metabarcoding might surpass traditional monitoring methods, for example through its ability to detect rare species, a key feature in early detection of NIS. Another interest of this approach is the identification of organisms difficult to identify based on morphology only (e.g., early developmental stages), making it relevant in the context of management programs. Because many marine benthic NIS have a bi-phasic bentho-pelagic life cycle, targeting their pelagic larval stages in zooplankton may allow early detection and assessment of their establishment and potential spread. We illustrate this approach with an analysis of bulk-DNA retrieved from a time-series of zooplankton samples collected over 22 months in one bay in Brittany (France). Using HTS of amplicons obtained with two markers (COI and 18S) and a metabarcoding approach, 12 NIS were identified and their temporal larval dynamics were monitored. Importantly, we chose to focus on a closed list of species, from four metazoan classes encompassing 52 NIS reported within the study area or nearby seas, with molecular references available or obtained locally for 42 of them. The use of a custom-designed database allowed the detection of three NIS otherwise not detected when using public databases. Interestingly, NIS known to have a short-lived larval stage were detected (e.g., the bryozoan Bugula neritina or the tunicate Corella eumyota). For two molluscs Ruditapes philippinarum and Crepidula fornicata, metabarcoding results were compared to those obtained using traditional methods (i.e., barcoding of individual larvae and morphology, respectively) to show the reliability of the approach in detecting and assessing the extent of their reproductive periods. Our results also revealed that the Pacific oyster Crassostrea gigas, a notorious invasive species, failed to reproduce in the study bay, showing that metabarcoding on larval stages also provides information regarding the establishment success (or failure) of NIS. While metabarcoding has its limitations and biases, this study demonstrates its effectiveness for surveillance of targeted NIS, notably to support management strategies like the European Marine Strategy Framework Directive (MSFD).