Nanopore sequencing technology offers an accessible form of genetic-based biomonitoring. However, this technology has not been assessed for utility in diatom-based bioassessment. We evaluated how DNA extraction method, primer choice, and bioinformatics treatment affect the results of diatom metabarcoding-based bioassessment using the MinION device. Samples were collected from streams in the Pacific Northwest, USA. DNA was extracted with two methods, amplified targeting both 18S and rbcL genes, and examined for diatom relative abundance with taxonomy-based and taxonomy-free bioinformatics methods. Diatom relative abundances generated from metabarcoding were compared to those generated from traditional microscopy. Taxonomic results varied significantly between metabarcoding and microscopy, with genetic alignments only achieving the level of genus or higher. The sequences generated using the 18S primer showed deeper taxonomic assignments than those generated with the rbcL primer, with a larger proportion of sequences only aligned to Phylum Bacillariophyta for the rbcL primer. The MinION produced highly reproducible data, with replicates of the same sample run with multiple barcodes on the same flow cell and on different flow cells showing highly similar relative abundances of the most abundant operational taxonomy units. In addition, Bray-Curtis dissimilarity between diatom communities from samples run multiple times on the same flow cell or on different flow cells was low, indicating high data reproducibility at the community level. Taxonomy-free OTU data were better able to distinguish samples from different streams and habitats than data aligned to databases. Both microscope and taxonomy-free metabarcoding produced similar bioassessment results, being able to separate the diatom community from low and high disturbance sites. While more work is needed for nanopore sequencing to achieve the same level of taxonomic diatom assessment as high throughput sequencing or microscopy methods, diatom-based bioassessment with the MinION device supports taxonomy-free analysis as an accessible and accurate tool for future biomonitoring studies.