A study was conducted to determine which subtests of the Gates Reading Readiness Tests have greatest predictive efficiency. An experiment to assess the effect of training in reading readiness on 222 kindergarten children (107 girls and 115 boys) provided the data for the research. The experimental group received 2 months' reading readiness training; the control group received none. Both groups were tested in April, June, and October. It was found that the ability can be based almost wholly "n the Letters and Numbers subtest. It is suggested that this one test be used to save time. It is also suggested that if similar tests are similarly analyzed, and those portions which provide redundant or slight information are eliminated, total testing time may be greatly reduced. (JB)