Recent theoretical results (Schlosser and Bailey, 1990; Schlosser et al., 1993) suggest that it is possible to model elements of metabolism to the extent necessary for flux control analysis using only the results of perturbation experiments. In particular, the development of detailed, mechanistic descriptions of certain relevant processes (which would otherwise require prohibitive effort) can be avoided by direct use of experimental results. Anaerobic glycolysis in Saccharomyces cerevisiae AMW-13C was studied both under unperturbed conditions and under several experimental perturbations. Changes in pathway fluxes and metabolite levels relative to the unperturbed values were observed and analyzed to determine the relationships that exist between several of these quantities. These relationships were used to refine and extend a previously determined model for the glycolytic pathway (Galazzo and Bailey, 1990, 1991). The refined model was then used to study the flux control characteristics of the pathway. The analysis indicates that the reaction catalyzed by phosphofructokinase (PFK), while retaining significant flux control, is less important than previous studies suggested and that glucose uptake is the predominant rate-controlling step when small changes are considered. The model indicates that larger increases in the amount of any single pathway enzyme yield no more than a 36% increase in ethanol production.