An approach to plant modeling that incorporates plant and environment interactions and that is driven by the requirements of designing and evaluating controlled ecological life support systems (CELSS) has been developed. The objective of this modeling approach is to permit the development of CELSS designs that optimize the performance of systems in which plant growth units are embedded.The approach described in this paper takes advantage of well-known analytical features of smoothly varying functions to construct an empirical model of the physiological response of plants to their environment. The model combines an emphasis on empirical data gathering to specify the plant response to nominal and near-nominal values of environmental control variables with general analytical relations that strictly hold for well-defined and behaved functions. Thus the same model is used for a variety of plant species and growing conditions, and numerical coefficients within the model define specific plant characteristics.The model development also suggests efficient experimental designs for collecting controlled environment plant response data. Examples of these experimental designs are also presented.