A battery of physical ability tests was validated using a predictive, criterion-related strategy. The battery was given to 1,003 female soldiers and 980 male soldiers before they had begun Army Basic Training. Criterion measures which represented physical competency in Basic Training (physical proficiency tests, sick call, profiles, and separation data) as well as on the job (lifting, carrying, pushing, pulling activities) were correlated with the soldiers' scores on the physical ability tests. The performance measures (criterion performance tasks) were designed to evaluate proficiency in the performance of tasks determined to be important in physically demanding Army jobs (Lift, Carry, Push and Torque). The criterion performance tasks were administered to the 951 soldiers who had completed Advanced Individual Training (AIT). Results indicated that test validity was high (R=.84). The Lift to 60 inches exercise accounted for 67 percent of the variance in criterion performance, while Lean Body Mass (LBM) and the Upright Pull test accounted for an additional 3 percent and 1 percent respectively. The fairness analysis showed that there were nonsignificant slope differences and only slight intercept differences which suggested minimal overprediction for women. (Author/JD)