DADA2 is a software package designed to model and correct errors in Illumina-sequenced amplicon data, inferring sample sequences with high resolution without coarse-graining into Operational Taxonomic Units (OTUs). It resolves differences as small as one nucleotide and has been shown to identify more real variants and fewer spurious sequences compared to other methods in mock communities. DADA2 was applied to vaginal samples from pregnant women, revealing previously undetected *Lactobacillus crispatus* variants. The software is open-source and available on GitHub, with package binaries through Bioconductor. DADA2 outperforms other methods in terms of accuracy and precision, allowing for the detection of fine-scale variation, which is crucial for understanding microbial community dynamics and health.DADA2 is a software package designed to model and correct errors in Illumina-sequenced amplicon data, inferring sample sequences with high resolution without coarse-graining into Operational Taxonomic Units (OTUs). It resolves differences as small as one nucleotide and has been shown to identify more real variants and fewer spurious sequences compared to other methods in mock communities. DADA2 was applied to vaginal samples from pregnant women, revealing previously undetected *Lactobacillus crispatus* variants. The software is open-source and available on GitHub, with package binaries through Bioconductor. DADA2 outperforms other methods in terms of accuracy and precision, allowing for the detection of fine-scale variation, which is crucial for understanding microbial community dynamics and health.