Marine autotrophic communities play a fundamental role in marine ecosystems, acting as primary producers and supporting entire food web. Understanding their structure, diversity, and distribution across different size fractions is essential for evaluating ecosystem functioning. In this study, we applied eDNA metabarcoding to investigate the taxonomic composition of plankton assemblages in coastal seawater samples, filtered through different size fractions: 100-50 µm, 50–10 µm, and 10–3 µm potentially corresponding to the size traits of phytoplankton. This approach enabled the characterization of micro- and nano-phytoplankton, including rare and or unculturable taxa. Seawater samples were collected from a temperate coastal site in the northwestern Adriatic Sea. Total genomic DNA was extracted from each size fraction and amplified using primers targeting the 18S V4 region of rRNA gene, followed by high-throughput sequencing. Preliminary results revealed distinct community profiles across the three size fractions, highlighting the ecological partitioning of taxa by cell size trait. At phylum level, the 100-50 µm fraction was dominated by large Alveolata (70%), including Dinoflagellata and Ciliophora subphyla, and by Stramenopiles (30%), mainly including the Bacillariophyta class of Mediophyceae, which consists of centric diatoms. In the 50–10 µm fraction, a slight shift was observed, with Stramenopiles (53%) more abundant than Alveolata (45%). The 10–3 µm fraction showed an enrichment in Stramenopiles (57%) and a reduction in Alveolata (27%). Additionally, Chlorophyta appeared in this fraction (15%), with small-sized families as Pycnococcaceae, Pyramimonadaceae, Chlorodendraceae. The size-structured distribution of these taxa reflected not only taxonomic diversity, but also functional and trophic differentiation, with important implications for coastal ecosystem functionality. The autotrophic-mixotrophic component was dominant in the smaller fractions, while the bigger one was dominated by the heterotrophic taxa. Thus, the eDNA metabarcoding with size-fractionated sampling can enhance the resolution of marine plankton community dynamics and enable a more detailed assessment of trophic structure and ecosystem functioning.