The purpose of this study was to compare the validity estimates of 10 psychologists, knowledgeable in both test content and job requirements, with validity coefficients obtained by correlating employees' test scores and a job performance criteria. Three departments of a large department store were used, Merchandise Clerical, Stock Records Clerical, and Supervisory Trainees. The correlations between test scores and merit ratings were generally higher for the two clerical departments than they were for the Supervisory Trainee department. The means of the psychologists' estimates of the test validities were not significantly correlated with the actual results for any of the three departments: .22 for the Merchandise Clerical department, .51 for the Stock Records Clerical department, and -.49 for the Supervisory Trainees. The validity estimates of each psychologist were then considered separately across all three departments. The correlation between estimated and actual validities was significant for only one psychologist.