The Luedeking-Piret model is an empirical relationship which is very widely used in cell cultures to evaluate specific production rates of some products (metabolites or others). It constitutes a very common method of calculation as much in fundamental as in applied research and especially for designing and optimizing industrial processes in very varied fields. However, this model appears to be frequently deficient and has to be greatly adapted, practically, one might say, for each individual case. Obviously, this is a very great drawback, requiring a great deal of time spent on it and one that greatly lessens the 'universality' of the model. This work reveals that it is possible to give the initial Luedeking-Piret model a much more general scope. The used method revealed metabolic switches that have never been suspected until now. Confirmation of the method would certainly give a precious general tool both to optimize production processes and to increase understanding of some physiological states of cells in chemostat.