This study analyzes selection in continuous culture as a means to improve the growth of microorganisms dependent upon the expression of extracytoplasmic enzymes. A quantitative, theoretical model was developed that considers increases in enzyme activity and/or expression due to mutation in conjunction with reaction and diffusion at the cell surface and in the surrounding boundary layer. This model was applied to a system consisting of a recombinant yeast cell growing on either soluble or insoluble substrates by virtue of extracytoplasmic enzymes either with or without tethering to the cell surface. Our results indicate that selection of faster-growing cells can be effective, arbitrarily defined as a faster-growing mutant representing 1% of the population in < or =3 months, but only under some conditions. For both soluble and insoluble substrates, tethering of enzymes to the cell surface is required for selection to be effective under the conditions examined. Significant increases in heterologous enzyme expression (2.5-fold for mutants as compared to the parent strain) are also required. In the soluble substrate/enzyme tethered case, the value of k(S) must also be low in order for selection to be effective. Cells growing on non-native substrates by virtue of extracytoplasmic enzyme production are expected to experience selective pressure in response to several additional factors, including cell shape, distance of the cell-substrate gap, properties of the gap, and perhaps mutation frequency. However, these factors exert a smaller impact on selection time and it is not clear that favorable values for these factors are required in order for selection to be effective.