The response of metabolic variables to small rate changes is quantitatively described by the control coefficients (Kacser & Burns, 1973; Heinrich & Rapoport, 1974). Owing to the existence of structural and kinetic constraints in metabolism, the control coefficients are not all independent. Two quantities are defined to evaluate how constrained metabolic control is: the fraction of independent control coefficients (f w ) and the number of independent control coefficients per independent variable [N(C/V)]. It is shown thatf w can be expressed in terms of the fraction of independent metabolite concentrations, the fraction of independent fluxes and the average fraction of independent metabolites affecting each rate.N(C/V) is equal to the average number of metabolite concentrations affecting each rate. The estimation of these quantities using experimental information available leads to the following conclusions concerning cellular metabolism: (i) only a small fraction of the control coefficients are independent; (ii) the number of rates (in average) that independently controls each independent variable is much smaller than its theoretical maximum; and (iii) the kinetic constraints are the main cause of the low value showed byf w andN(C/V). Finally, some arguments are given, that could explain why living organisms do not evolve to less constrained metabolic responses.