This study sought to define the relationships between reading skill, reading requirements, and success in a sample of navy Class "A" schools. A further objective was to provide a methodological demonstration of the procedures required to allow an inference about causality in these relationships. Reading skill was compared with a nonverbal measure of general ability and scores on the navy's Basic Test Battery as predictors of school success and failure. Results showed wide variations in the dependence of performance on reading skills among the schools, as well as the amount and difficulty of the reading they require. Reading skill and general ability were as good as or better than course selector tests as predictors of school performance in some schools. A discussion of the advantages and disadvantages of several options for dealing with deficient reading is included. (Author)