The study was designed to probe the changes that might occur in achievement test results if test items were made easier to read. The rewording of the test, the cross-validation of items by a panel, and the application of readability formulas to the reworded test gave, in theory, a comparable instrument. By agreement of readability formulas (Dale and Chall, SRA, Fry), the reworded test was easier to read. Of the 412 ninth grade students, 198 were given the revised Florida State-Wide Testing Program and 214 were given the reconstructed test. All were given the Comprehensive Test of Basic Skills. The findings seemed to indicate that making a test easier to read did not produce changes in the results of the test. (Author/EB)