The research included a critical incident job analysis to develop a behaviorally anchored rating scale. The officers were administered the IS3, and after 6 weeks, performance data were collected. Criteria validation revealed that the instrument accounted for about 20 percent of the variance in absence, lateness, and supervisory ratings. No significant variance was found for supervisors' overall assessments of officers. Discriminant analysis indicated that the IS3 was better at classifying persons on a performance appraisal rating (acceptable versus unacceptable) and potential for absenteeism and lateness than it was at predicting the termination and employment decisions of supervisors. Although raters appeared to display a significant race bias in evaluating black officers, the IS3 was not found to adversely affect these officers. The IS3's performance in this study merits consideration in future research with larger sample sizes and uncontaminated criteria. Tables and 9 references (Author abstract modified)