The high therapeutic and financial value offered by polyclonal antibodies and their fragments has prompted extensive commercialization for the treatment of a wide range of acute clinical indications. Large-scale manufacture typically includes antibody-specific chromatography steps that employ custom-made affinity matrices to separate product-specific IgG from the remainder of the contaminating antibody repertoire. The high cost of such matrices necessitates efficient process design in order to maximize their economic potential. Techniques that identify the most suitable operating conditions for achieving desired levels of manufacturing performance are therefore of significant utility. This paper describes the development of a computer model that incorporates the effects of capacity changes over consecutive chromatographic operational cycles in order to identify combinations of protein load and loading flowrate that satisfy preset constraints of product yield and throughput. The method is illustrated by application to the manufacture of DigiFab, an FDA-approved polyclonal antibody fragment purified from ovine serum, which is used to treat digoxin toxicity (Protherics U.K. Limited). The model was populated with data obtained from scale-down experimental studies of the commercial-scale affinity purification step, which correlated measured changes in matrix capacity with the total protein load and number of resin re-uses. To enable a tradeoff between yield and throughput, output values were integrated together into a single metric by multi-attribute decision-making techniques to identify the most suitable flowrate and feed concentration required for achieving target levels of DigiFab yield and throughput. Results indicated that reducing the flowrate by 70% (from the current level) and using a protein load at the midpoint of the range currently employed at production scale (similar to 200-500 g/L) would provide the most satisfactory tradeoff between yield and throughput.