Whooping cough vaccines are produced using different ranges of cultivation conditions and medium compositions, which are known to influence growth rate, virulence factor production and degradation, as well as the virulence factors' association to the cell. This study quantifies the impact of individual parameters on Pertussis Toxin (PT) production, using an optimized chemically defined medium as starting point, rather than a complex medium. A number of chemicals that are identified affect both growth rate and virulence factor production, which occur at similar levels in various commonly used production media. Also, degradation by proteolytic activity is shown to be an important parameter to monitor, since it significantly affects the PT yield. Low sodium concentrations, i.e. 50-75mM rather than the conventional 100-140mM, significantly increase the growth rate of the organism, the final optical density, as well as the association of PT to the cells. The absolute amount of biomass produced measured as dry weight, is similar for all sodium concentrations tested, contrary to earlier work. While it is known that high iron concentrations inhibit virulence factor production, it is shown here that iron-limited growth results in very high specific PT production. This finding may be used to produce a whole-cell vaccine with little biomass per dose, reducing whole-cell vaccine toxicity. The Bordetella pertussis strain 509 used here produces 30% more PT at 34 than at 37 degrees C, a commonly used cultivation temperature. The data in this study show that existing production processes for cellular and acellular vaccines can in principle be optimised considerably by taking simple measures.