The Phonic Inventories are a three level spelling test which was developed to enable identification of the patterns of spelling error made by children, as an aid for teachers and therapists. The aim of this study is to test the assumptions of face validity, content validity, construct validity and discriminant validity of the Phonic Inventories, as well as their short and long-term reliability. This has been done by using both longitudinal and cross-sectional datasets from 1979 and 2001. The results of the study indicate that the three levels of the instrument are reliable, and that they have potential for clinical as well as classroom use in determining which alphabetic rules learners have or have not acquired. They also have potential for use by teachers for screening purposes, with the aim of identifying learners experiencing difficulties in learning the rule systems used in reading and writing.