Background: Oral disease pathogenesis is primarily driven by microbial dysbiosis although diagnosis is routinely macroscopic. To improve early detection especially in HIV patients who are disproportionately affected, there is need to reconcile macroscopic and microscopic characteristics of disease. This study aimed to use amplicon sequencing data to characterize oral microbiota changes along the decayed, missing, filled teeth (DMFT) index.Methods: Amplicon sequencing of the V6-V8 region of the 16S rRNA gene was done on DNA recovered from whole unstimulated saliva of 59 HIV positive and 29 HIV negative individuals, respectively. The microbial structure, composition and co-occurrence networks were characterized using QIIME-2, Phyloseq, Microbiome-1.9.2 and Metacoder in R.Results: From a total of 2.6 million high quality sequence reads obtained, we characterized the oral microbiota into 2,093 operational taxonomic units (OTUs), 21 phyla and 239 genera. While oral microbiota did not cluster participants into distinct groups that track with the DMFT index, we observed the following: a) A steady increase in accessory microbiota while the core size (~50% of richness) remained stable. b) The abundance of core genera such as Stomatobaculum, Peptostreptococcus and Campylobacter increased at onset of pathology (low DMFT), c) A general increase in oral microbial biomass with a typical log difference between gingivitis and periodontitis. d) The onset of pathology (low DMFT) was associated with massive reduction in oral microbial entropy.Conclusions: Although oral microbial shifts along the DMFT index are not distinct, we have demonstrated the potential utility of microbiota dynamics to inform oral disease characteristics. We therefore propose a framework to inform future clinical oral metagenomic studies especially among HIV positive persons in resource limited settings.