The study's hypothesis was that the PSI-5 dishonesty attitude scale would be a reliable predictor of employee theft. The studies also explored how other attitudinal and behavioral measures from the PSI-5 would correlate with employee theft. Employees from one store were also given a locus of control scale and a social desirability scale to determine personality correlates of employee theft. One study involved 27 employees of a store in Chicago while the other included 15 employees of a store in Cincinnati. Results showed that 56 percent of the Chicago store employees and 40 percent of the Cincinnati store employees made employee theft admissions. The average value of cash and merchandise stolen within the last 3 years was $46.70 per Chicago employee and $95.83 per Cincinnati employee. For both sets of employees, the dishonesty attitude score was positively and significantly associated with total employee theft. This finding strongly supported the study's main hypothesis. For both stores, the major consistent predictors of employee theft were, in order, the dishonesty attitude score, the total dollar amount of nonemployee theft, and the amount of alcohol consumed by a person in a typical week. The personality variables studied were not correlated with employee theft. Results indicated that test measures can be used to identify and screen the potential employee thief and thus reduce financial losses to businesses. It was concluded that personnel selection is a legitimate concern for security administrators and should not be the sole responsibility of the personnel departments. Personnel department staff need to be educated about employee theft. Footnotes with references and appendixes presenting data are included.