The Laurentian Great Lakes are a vast interconnected freshwater system that provides essential ecosystem services to tens of millions of people. It spans large gradients of temperature and productivity and is experiencing rapid environmental change. Microbial communities are at the heart of this system, playing critical roles in biogeochemical cycling and food webs, while also serving as sentinels of change. This dissertation investigates microbial community structure, diversity, and dynamics across the Laurentian Great Lakes using the longest (2012–2019) 16S rRNA gene sequencing dataset collected to date. This dataset captures prokaryotes and chloroplast-containing eukaryotes across two size fractions. This unprecedented dataset enabled a basin-wide, size-resolved analysis of microbial dynamics and their environmental drivers across space and time in one of the world's fastest-warming freshwater ecosystems. Chapter 1 examines prokaryotic communities, showing that thermal stratification and lakebed bathymetry drive patterns in diversity. Our results show that surface communities become less diverse as stratification persists, and deep water communities converge across large distances at the deepest stations. Chapter 2 explores the first Great Lakes-wide molecular survey of chloroplast-containing phytoplankton. This dataset reveals that large-cell communities are relatively stable over the timeseries, but picophyoplankton show increasing variability and expansion in the upper lakes, likely due to invasive mussels and nutrient shifts. Chapter 3 uses Latent Dirichlet Allocation (LDA) modeling to identify microbial sub-communities that respond coherently to environmental conditions. This new approach revealed co-responding sub-communities across size fractions and kingdoms. Sub-communities of free-living prokaryotes tended to mix across locations and overlap in composition, while particle-associated and picophytoplankton sub-communities were more taxonomically distinct and locally endemic. Together, this work establishes the foundations for understanding microbial biogeography, ecological specialization, and size-fractionation dynamics in the Great Lakes. It also demonstrates the power of combining long-term datasets, ecological modeling, and size-resolved analyses to anticipate how microbial communities may respond to ongoing environmental change.