A growing focus in microbial ecology is understanding how beneficial microbiome function is created and maintained through various assembly mechanisms. To date, most studies focus on the contemporary factors of a local community finding that a microbe’s niche requirements drive species distributions in a deterministic manner. However, historical factors, such as dispersal limitation, drift, and priority effects, likely exert a large influence on the hyper-diverse metacommunity of host-associated microbes. These historical factors likely explain the high dissimilarity often found between the microbiomes of conspecifics under identical environmental conditions. Thus, to fully harness the potential of microbiome-based therapies, we need to incorporate historically contingent interactions into our models. This talk explores these themes in two model systems: amphibian and plant microbiomes, using both observational and manipulative experiments. It also explores the use of metagenomics and comparative genomics in community ecology to tease apart historical contingencies driving structure and function at the level of species and strains.