Screening for microbiome modulators requires availability of a high-throughput in vitro model that replicates subgingival dysbiosis and normobiosis, with a tool to measure microbial dysbiosis. Here, we tested various formulations to grow health- and periodontitis-associated subgingival microbiomes in parallel, and we describe a new subgingival dysbiosis index. Subgingival plaque samples pooled from 5 healthy subjects and, separately, 5 subjects with periodontitis were used to inoculate a Calgary Biofilm Device containing saliva-conditioned, hydroxyapatite-coated pegs. Microbiomes were grown for 7 d on either nutrient-rich media—including a modification of SHI medium, brain-heart infusion (BHI) supplemented with hemin and vitamin K, and a blend of SHI and BHI, each at 3 sucrose concentrations (0%, 0.05% and 0.1%)—or nutrient-limited media (saliva with 5%, 10%, or 20% inactivated human serum). The microbiomes were assessed for biomass, viability, and 16S rRNA profiles. In addition to richness and diversity, a dysbiosis index was calculated as the ratio of the sum of relative abundances of disease-associated species to that of health-associated species. The supplemented BHI and blend of SHI and BHI resulted in the highest biomass, whereas saliva-serum maximized viability. Distinct groups of bacteria were enriched in the different media. Regardless of medium type, the periodontitis-derived microbiomes showed higher species richness and alpha diversity and clustered with their inoculum separate from the health-derived microbiomes. Microbiomes grown in saliva-serum showed the highest species richness and the highest similarity to the clinical inocula in both health and disease. However, inclusion of serum reduced alpha diversity and increased dysbiosis in healthy microbiomes in a dose-dependent manner, mainly due to overenrichment of Porphyromonas species. The modification of SHI stood second in terms of species richness and diversity but resulted in low biomass and viability and significantly worsened dysbiosis in the periodontitis-derived microbiomes. Overall, saliva with 5% human serum was optimal for replicating subgingival microbiomes from health and disease.