Summary:Anumberof personnel managers who have adopted selective test procedures have been surprised and even alarmed to discover that within a short time almost all applicants qualify on the test batteries, even though the installations were carefully made and flexible selection standards established. No great problem arises, for the employment manager can effectively adjust his critical scores to meet the fluctuation of the labor market; however, the reasons for the superiority in the test level of new applicants as compared to the original criterion groups present an interesting problem to be considered. Although the answers are not clear‐cut, the implications for such changes in qualifications are the subject of this report. These indicate that the existence of a testing program attracts the better applicants and discourages the poorerA selective testing program was installed for all productive workers in three plants of the same industry. One plant was in Illinois, one in Maryland, and the third in Washington, D. C. All three plants are members of the same national association. This is a borderline industry, employing womenSelection tests were validated by testing present employees in all three plants. Subsequent applicants were superior to the criterion groups to such a degree that initial standards failed to screen out the expected number of applicants. The differences between the experimental groups and subsequent applicant groups are statistically significant. These differences cannot be explained in terms of shifts in the labor market or motivation. It is suggested that self‐imposed selection may take place within the labor source, and that only the more capable applicants apply at plants where it has become known that test qualifications have been made a requisite for employm