Performance data on 386 male correction officers were used to validate the IPI; the cross-validation sample consisted of 293 officers. Predictions for poor performance using the IPI were up to 41 percent more accurate than predictions made without the psychological and biographical information. While the findings do indicate the utility of a behaviorally focused screening instrument, the validity results are likely to be underestimates of true validities. Identification may have been truncated because of range restriction, criterion fallibility, or sample idiosyncrasies. 2 tables and 33 references