A low—resolution simulation model of transpiration was developed and run, with data from field studies in southwestern Oregon. The output of the model served as a means of relating environmental variables to plant response. This relation was used to define an ordinate which, in conjunction with previously developed ecosystem ordinates, proved helpful for comparing ecosystems, predicting community composition and, in special cases, growth. The data requirements of the model are modest; we deliberately developed a model that can be used on data obtained from field studies where electric power is unavailable and use of sophisticated instrumentation is impossible. The model requires inputs of air and soil temperature, atmospheric humidity, seasonal plant water potential (expressed as plant moisture stress, the absolute value of plant water potential), and a model of stomatal behavior. Where it was impossible to obtain accurate data, stochastic models were used to provide the necessary input. The model simulated both potential and actual transpiration, the ratio of which is the most valuable single index of the seasonal moisture regime. Where no measurable stomatal control was exerted by Douglas—fir, the ratio was 1.0, indicating that adequate water was available to meet the transpiration demand. The ratio approached 0.3 on the drier locations. Significant changes in vegetation and growth were associated with this index.