DNA-barcoding utilises a fragment of the mitochondrial cytochrome oxidase subunit 1 (COI) gene to identify most animal species. Using next generation sequencing (NGS), this method can be further developed into metabarcoding processes that allow the simultaneous identification of several species from a mixed sample. We created a database of COI sequences of 27 mosquito species collected in Sweden, and combined our data with 27 additional sequences from GenBank to cover the taxa recently documented in Sweden and to include possible invasive taxa. Comparisons show that COI metabarcoding reliably identifies 41 of 54 species and the remainder to species group. Using three independent primer pairs along the COI gene, we further developed this barcoding approach to simultaneously identify Swedish mosquitoes in communities using NGS and quantify relative abundance of each mosquito species in the sample, using bioinformatics methods. We tested the accuracy of the metabarcoding method using communities assembled from morphologically identified mosquitoes, revealing 80% positive identification rate and the estimates of population structure which reflects the input sample. We conclude that metabarcoding is useful as a high throughput identification technique and for the quantification of species.