For many metabolic systems, available exptl. data allow description of the system by elasticities and control coeffs. The availability of information of this kind motivated the development of a (log)linear kinetic model of metabolic systems that is completely and explicitly detd. by this information. It is shown here that this model can accurately describe the dynamic responses of metabolic systems that exhibit strong non-linearities. Based on the excellent approxn. provided by the (log)linear model, a method is developed for the estn. of the performance of metabolic systems subject to spatiotemporal variations of the system parameters and the process operating conditions. The method suggests expts. that can quantify the effect of these variations. Study of a model glycolytic pathway illustrates the applicability and the usefulness of this framework. Time-av. flux control coeffs. are shown to vary strongly and not monotonically as the period of the external variations changes. [on SciFinder (R)]