Cell surface display of polypeptide libraries combined with flow cytometric cell sorting presents remarkable potential for enhancement of protein-ligand recognition properties. To maximize the utility of this approach, screening and purification conditions must be optimized to take full advantage of the quantitative feature of this technique. In particular, discrimination of improved library mutants from an excess of wild-type polypeptides is dependent upon an effective screening methodology. Fluorescence discrimination profiles for improved library mutants were derived from a mathematical model of expected cell fluorescence intensities for polypeptide libraries screened with fluorescent ligand. Profiles for surface-displayed libraries under equilibrium or kinetic screening conditions demonstrate distinct discrimination optima from which optimal equilibrium and kinetic screening parameters were derived. In addition, a statistical model of low cytometrically analyzed cell populations indicates the importance of low-stringency sorting followed by amplification through regrowth and resorting at increased stringency. This analysis further yields quantitative recommendations for cell-sorting stringency.