Characterization of flocculation for cell removal from fermentation broth via polyelectrolyte addition is commonly based on qualitative methods such as physical appearance of the floc. The use of zeta potential as a quantitative measure of floc character was evaluated as an indicator of optimal polymer addition. Zeta potential was found to increase with increasing polyelectrolyte dosage, but never reached zero indicating flocculation occurs at least partially through a bridging mechanism. Experiments were conducted using various polymer concentrations (25-75g/L) and dosing methods (batch, incremental and continuous addition) that resulted in variable overall polymer requirements to achieve optimum flocculation. Zeta potential was found to be constant at optimal floc character regardless of the total amount of polymer added, polymer concentration, or method of polymer addition. Polymer requirements to achieve a particular floc character can vary greatly depending on polymer dosing conditions and fermentation batch. Polymer dosing method and calcium concentration were both found to have a significant effect (p<0.0001) with continuous polymer addition and high calcium concentration requiring less polymer than did batch polymer addition and low calcium concentration. Polymer dosing concentration did not significantly affect polymer requirement for optimal flocculation. The effectiveness of polyelectrolyte addition to fermentation broth for cell flocculation without extracellular enzyme loss was studied. Partitioning of the enzyme subtilisin between the liquid and the sedimented Bacillus floc was the focus. Enzyme partitioning to the floc increased as the polymer dosage was increased beyond that necessary to obtain optimum floc character (brain floc) for cell removal by centrifugation. Partitioning to the cell floc at high polymer dosages was determined to be at least partially a result of a direct interaction between the polyelectrolyte and enzyme, and not only a result of physical entrapment in the cell floc matrix. Enzyme loss was not likely due to pH denaturation during the flocculation process. The direct interaction between polyelectrolyte and enzyme was characterized through turbidimetric titrations and partitioning studies. Any reductions in enzyme loss at optimum flocculation for cell removal by changing the polymer feed concentration or the method of polymer addition were not discernable.