The objective of the present research was to investigate the effects of test-wiseness training on achievement. The study involved 208 Black, White and Hispanic students in grades three and five. An experimental training session was employed to test whether four aspects of test-wiseness (following directions, guessing, answer changing, and use of time) would interact with race and sex in math and reading achievement using a post- and delayed post-test. For the fifth grade, the results indicated that students in the training groups scored higher than the controls in math on both the post- and delayed post-test. No training effects were significant at the third grade, nor were any of the interactions between training and race or training and sex at either the third or fifth grade. Recommendations for futher research are offered. (Author)