Although the development of limited-area, regional-scale numerical-weather prediction models has been driven largely by the practical need to improve quantitative precipitation forecasts, a related motivation has been scientific interest in investigating mesoscale phenomena and processes under “controlled” conditions. The establishment of mesoscale meteorology as a recognized discipline of the atmospheric sciences provides a context for exploring how regional models can be used effectively as research tools. It is argued that these models are sufficiently advanced and refined to provide investigators with four-dimensional, dynamically consistent data sets to supplement and extend those available from observations. Consequently, regional models combined with observations potentially are of considerable utility in conducting case studies of mesoscale phenomena. These contentions are illustrated with examples from the literature and with suggestions for future research involving the application of regional models to cast studies of midlatitude mesoscale phenomena associated with significant precipitation or severe weather. Perceptions of how the advent of regional models is contributing to fundamental changes in traditional synoptic meteorology are discussed.