Data collected in controlled-environment (CE) and free-air carbon dioxide enrichment (FACE) facilities have different strengths relative to use in developing crop growth models.  For example, CE are the only facilities that allow studying temperature effects on basic processes, but they are less reliable for absolute biomass and yield because of single-plant, potted plant, and bordering issues.  On the other hand, FACE facilities do not allow studying temperature responses, but do reflect field soil and climatic conditions and should give better control of border effects on absolute yields.  The CE facilities are relied on for temperature effects on rate of leaf appearance, rate of reproductive development (phenology), rate of leaf photosynthesis, rate and duration of single seed growth, although light levels near half full sunlight are needed.  Certain parameters such as quantum efficiency are so well researched that no one need measure them again.  Time-series measurements of land-area-dependent canopy assimilation, biomass, growth, and yield are only as good as the extent of bordering to account for plants in full crop stands, whether studied in CE or FACE.  For this reason, non-bordered single plant data are not relied on, in an absolute sense, to set crop model productivity parameters.  Single-plant data has greater variability than large-area samples, regardless of potted plant, CE or FACE.  Furthermore, none of these data sources should be used in absence of full testing of crop growth models with large land-area samples of growth and yield collected from field studies in a diverse range of climatic conditions.  FACE sites are too few in number to fulfill the latter requirement.