The human oral microbial ecosystem is one of the best characterized and highly complex bacterial communities known. It is estimated that about 600 bacterial species exist in the mouth and that 85% of those are currently known at the molecular (16 S rRNA) level. The major bacterial physiologies occurring in the oral cavity have been known for many years, and knowledge exists of the distribution of organisms in time (as plaque accumulates) and in space (different environments within the mouth). However, only rudimentary data are available on interactions between the bacterial species. It is precisely these interactions that, along with the interactions of the developing community with its human host, must drive the succession of genera that is observed to occur. In the reductionist scientific approach to studying such interactions, disrupted plaque is used to isolate single organisms, and interactions between these organisms are examined by recombining the organisms ex situ. Lessons learned from these in vitro studies can be applied to understand empirical observations made in vivo. However, this approach begins with the primary assumption that the chosen interaction does in fact occur in vivo. The accessibility and the well-characterized nature of the oral ecosystem presents an opportunity for approaching the problem from the opposite direction; one can capture a community very early in development in vivo, then apply in vitro methods to sort out the interactions within that community. This latter approach begins with a set of organisms known to interact in vivo. A combination of both approaches should yield robust microbiological data suitable for in silico modeling and analyses.