This article provides a method for estimating the utility of a choice between alternative predictors for selecting a single cohort of new employees when some of these employees are later judged to perform at an unacceptable level and are therefore dismissed at the end of the probationary period. The method is shown to apply to both the traditional selection utility situation, in which the objective is to obtain a fixed number of hires, and the situation in which the objective is to obtain a fixed number of employees who successfully survive the initial probationary period. The utility estimation method proposed applies J. W. Boudreau and C. J. Berger's (1985) conceptual model to a singlecohort probationary case and extends that model by providing an estimation method to predict the utility of alternative predictors before selection takes place. The limitations of the proposal as well as its relationship to recruitment and separation costs are also discussed. I thank William Whitely and two particularly helpful anonymous reviewers for their useful comments on a draft of the article.A computer program (written in FORTRAN 77) that performs the utility calculations described in this article is available on request. Because the program calls routines from the Numerical Algorithms Group library, interested readers should first check whether they have this library (or a similar one) at their disposal.