Carefully designed experimental measurement programs are necessary to the task of collecting the extensive data base needed for testing air pollution models. In this paper a series of recommendations is offered as input for field measurement programs. These recommendations update and refine suggestions made earlier. Some specific techniques and instrumentation are called for in certain cases, but in others only conceptual objectives are set forth. The report presents three specific conceptual plans. A description is presented of a data management system using the Los Angeles Reactive Pollutant Program (LARPP) as a prototype. It is concluded that future field programs must be structured to meet specific goals. The specific goals discussed here include verifying emissions models for vehicle populations, following photochemical transformations in a moving air mass, searching for turbulent reaction inhomogeneities in the atmosphere, and characterizing the interaction of gas phase species with aerosols. (Modified author abstract)