Describes paper and pencil predictions of employee theft and examines studies of validity, reliability, and adverse impact of these tests. Results showed consistently positive correlations, but identified a variety of methodological differences which make the direct comparison of test validities suspect. (LLL)