This article presents the extension of the one-cohort probationary selection (OCPS) framework (De Corte, 1994) to a multiple cohort model and shows how the extension can be applied to estimate the utility of probationary selection with replacement (PSWR) when a ®xed quota of successfully performing selectees is required. Next, formulas to compute the net present value of the OCPS and the PSWR approach to the selection of a ®xed quota of successful employees are derived. The equations indicate that the latter approach will usually lead to a higher utility, but not necessarily to a higher net present value of the predictor selected workforce. This ®nding, which is corroborated by the results of an example application, is of immediate practical relevance because it justi®es the probationary selection with replacement practice that is adopted in many West European countries. Finally, the example results suggest also that the traditional Brogden±Cronbach±Gleser framework generally overestimates the incremental utility of using a predictor when the objective is to select a ®xed quota of successful selectees. *