Experimental and model studies have been performed to characterize the control properties of hexokinase and phosphofructokinase in muscle glycolysis and to examine the nature of error associated with experimental flux control coefficient determinations. Different approaches of metabolic control analysis, classical titration, co-response analysis and kinetic modelling indicated that flux control coefficients could be reliably estimated experimentally for the upper part of glycolysis. The kinetic parameters applied to construct the mathematical model were determined in muscle extract under similar conditions used for flux studies. If the kinetic parameters of commercial enzymes are introduced into the model the control analysis data cannot be trusted. Co-response analysis can also be successfully applied to determination of the flux control coefficients of the system. However, the involvement of a rapid-equilibrium enzyme, such as glucose 6-phosphate isomerase, could result in estimation errors for the relevant co-response coefficients that are propagated into the elasticity matrix. If the co-response coefficients related to isomerase activity are replaced by the values obtained by kinetic modelling, the values of elasticities are correct. Our data also suggest that in the upper part of glycolysis hexokinase mainly controls the pathway flux whereas phosphofructokinase exerts dominant control on the turnover of internal metabolite stocks inside the system.