Vibrio gazogenes ATCC 29988 growth and prodigiosin synthesis were studied in batch culture on complex and defined media and in chemostat cultures on defined medium. In batch culture on complex medium, a maximum growth rate of 0.75 h and a maximum prodigiosin concentration of 80 ng of prodigiosin . mg of cell protein were observed. In batch culture on defined medium, maximum growth rates were lower (maximum growth rate, 0.40 h), and maximum prodigiosin concentrations were higher (1,500 ng . mg of protein). In batch culture on either complex or defined medium, growth was characterized by a period of logarithmic growth followed by a period of linear growth; on either medium, prodigiosin biosynthesis was maximum during linear growth. In batch culture on defined medium, the initial concentration of glucose optimal for growth and pigment production was 3.0%; higher levels of glucose suppressed synthesis of the pigment. V. gazogenes had an absolute requirement for Na; optimal growth occurred in the presence of 100 mM NaCl. Increases in the concentration of Na up to 600 mM resulted in further increases in the concentration of pigment in the broth. Prodigiosin was synthesized at a maximum level in the presence of inorganic phosphate concentrations suboptimal for growth. Concentrations of KH(2)PO(4) above 0.4 mM caused decreased pigment synthesis, whereas maximum cell growth occurred at 1.0 mM. Optimal growth and pigment production occurred in the presence of 8 to 16 mg of ferric ion . liter, with higher concentrations proving inhibitory to both growth and pigment production. Both growth and pigment production were found to decrease with increased concentrations of p-aminobenzoic acid. The highest specific concentration of prodigiosin (3,480 ng . mg protein) was observed in chemostat cultures at a dilution rate of 0.057 h. The specific rate of prodigiosin production at this dilution rate was approximately 80% greater than that observed in batch culture on defined medium. At dilution rates greater than 0.057 h, the concentration of cells decreased with increasing dilution rate, resulting in a profile comparable to that expected for linear growth kinetics. No explanation could be found for the linear growth profiles obtained for both batch and chemostat cultures.