Difficulties with in vivo studies of natural plaque and its complex, heterogeneous structure have led to development of laboratory biofilm plaque model systems. Technologies for their culture are outlined, and the rationale, strengths, and relative uses of two complementary approaches to microbial models with a focus on plaque biodiversity are analyzed. Construction of synthetic consortia biofilms of major plaque species has established a variety of bacterial interactions important in plaque development. In particular, the 'Marsh' nine-species biofilm consortia systems are powerful quasi steady-state models which can be closely specified, modified, and analyzed. In the second approach, microcosm plaque biofilms are evolved in vitro from the natural oral microflora to the laboratory model most closely related to plaque in vivo. Functionally reproducible microcosm plaques are attainable with a biodiverse microbiota, heterogeneous structure, and pH behavior consistent with those of natural plaque. The resting pH can be controlled by urea supply. Their growth patterns, pH gradient formation, control of urease levels by environmental effectors, and plaque mineralization have been investigated. Microcosm biofilms may be the only useful in vitro systems where the identity of the microbes and processes involved is uncertain. Together, these two approaches begin to capture the complexity of plaque biofilm development, ecology, behavior, and pathology. They facilitate hypothesis testing across almost the whole range of plaque biology and the investigation of antiplaque procedures yielding accurate predictions of plaque behavior in vivo.