Molecular evolutionary analyses commonly focus on individual genes and their products without considering functional interactions necessary for organismal fitness. There is a need to understand how evolutionary forces act on multiple interacting genes together that function as components of molecular pathways and networks. A question that emerges is how stable rate-limiting steps are over evolutionary time? Although the biochemistry community has generally assumed that metabolic pathways have evolutionarily stable rate limiting steps, the idea has not been rigorously tested. Glycolysis is an essential metabolic pathway that involves a breakdown of one glucose molecule to two pyruvate molecules. Common thought is that control in glycolysis is governed by reaction free energy and pathway position. The prevailing view is that the rates of individual enzymes in glycolysis are controlled by reaction free energy and their relative positions in the pathway. Here we performed a sensitivity analysis in all available glycolysis kinetic models and discovered that rate-limiting steps varied between different species. Another approach to evaluate evolutionary instability of rate-limiting steps is to use an evolutionary computational simulation of a glycolysis-like synthetic pathway with mutation and various selective schemes. The simulations performed used combinations of selection for steady state flux. That were compared against the cost of synthesizing mRNA and proteins, and against the accumulation of high concentrations of a deleterious intermediate. Results from this study suggest that rate-limiting steps in metabolic pathways are generally not evolutionarily stable and are subject to co-evolutionary mutation-selection balance. We further hypothesized that when the system is out of equilibrium due to fluctuating selection, population size or due to positive directional selection, the evolutionary stability of rate limiting steps will have a different evolutionary dynamic. Depending upon the underlying population genetic regime, fluctuating population size was found to increase the evolutionary stability of rate limiting steps in some scenarios. This result suggests that differences in patterns of evolution when systems are in and out of equilibrium, may lead to predictable differences in observed patterns for divergent evolutionary scenarios.