DNA metabarcoding can be a promising alternative to microscopy for analysing phytoplankton, a key ecological indicator for freshwater ecosystems. The aim of this study was to evaluate the performance of different barcodes and associated primer pairs to assess microalgal diversity with DNA metabarcoding using a single barcode targeting all microalgae. We investigated barcodes in 16S and 23S rRNA genes, encoding for prokaryotic ribosomal sub-units, that are present in Cyanobacteria as well as in chloroplasts. In silico PCR tests were carried out on eight 16S and five 23S primer pairs using the Phytool reference library. Two and three pairs were selected for 16S and 23S, respectively, to perform an in vitro metabarcoding test based on a mock community made of DNA extracts of 10 microalgae strains. The 23S pairs enabled to detect all species, whereas 16S ones failed in the detection of some of them. One pair was selected for each genetic marker, based on its efficiency and specificity towards microalgae (e.g. not heterotrophic bacteria). Another mock community covering a larger diversity (18 microalgae strains) was used to test the efficiency of the selected pairs and their ability to estimate relative abundances. The 23S pair performed better than the 16S one for detecting target species with also more accuracy to assess their relative abundances. We conclude that the 23S primer pair ECLA23S_F1/ECLA23S_R1 appears as a good candidate to decipher freshwater phytoplankton communities. As a next step, it will be necessary to confirm these results on a large diversity of natural communities.