Bacterial endosymbiosis is widespread among Bivalvia. Symbiosis between lucinid bivalves and sulfur-oxidizing bacteria has received recent attention, as lucinids are one of the geologically oldest extant bivalve clades to possess endosymbionts. However, the ecological and evolutionary relationships between host and symbiont are poorly understood. Reconstructing the evolutionary history and geologic significance of lucinid endosymbiosis will be difficult until endosymbiont ecology and taxonomic diversity are better characterized. Our goal was to uncover lucinid endosymbiont diversity within a single modern habitat and to compare endosymbiont diversity to the full bacterial diversity of that site. Host organisms (Lucinisca nassula and Lucina pectinata) were collected from Cedar Keys, Florida, at ~15-20 cm sediment depth in sea grass beds. PCR amplification and sequencing of bacterial 16S rRNA genes from sediment cores and lucinid gills retrieved ~5200 sequences. Sediment diversity was high, represented by 13 major taxonomic groups, including equally dominant Chloroflexi, δ-, and γ-proteobacteria. Other organisms included the Bacteroides, Acidobacteria, Spirochaetes, and Firmicutes. Previous work suggested that symbionts may exist outside of the host, and we retrieved rare (<2%) sequences related to gill symbionts from the sediment. Based on comparative phylogenetic methods, gill sequences were most closely related to uncultured γ-proteobacteria associated with symbiosis, and specifically to lucinid endosymbionts (97-99% sequence similarity) and not to free-living organisms. Interestingly, not all gill sequences were genetically identical, with intra- and inter-gill sequence diversity. Our results support the hypothesis that recruitment of free-living organisms is likely. But, based on habitat geochemistry and assumed metabolic preferences, the bacteria are constrained to reducing conditions, and this may be reflected in the habitat types colonized by the host. However, as lucinids are found in geochemically variable and geographically separated habitats, we are testing habitat-host-symbiont diversity from other locations. These results will be significant to paleoecological and evolutionary studies using lucinids in the fossil record (e.g. isotope studies).