The validity of inferences drawn from readiness tests used to place children in a regular or a 2-year kindergarten program was investigated. Data from nine school districts in Virginia were used in the study. The study explored technical characteristics of four commonly used readiness tests: the Brigance Kindergarten and First Grade Screen, the Developmental Indicators for the Assessment of Learning-Revised (DIALR), the Daberon Screening for School Readiness, and the Missouri Kindergarten Inventory of Developmental Skills (KIDS). Findings indicated that male, minority, low socioeconomic status, and young children scored consistently lower than others on all four tests. Three of the tests were fairly reliable, but the KIDS was highly inconsistent for minorities and those under 5 years of age. For these children, true scores could be nearly 20 percent higher or lower than those obtained. Although the Brigance, Daberon, and DIALR scores were more consistent indicators than anticipated, they fell short of standards which recommend reliability coefficients of .90 or higher for placement decisions. No test was an impressive predictor of future test performance. Findings are interpreted in the context of specific selection and placement policies. (Author/RH)