This study describes the writing, refining, and field testing of six experimental reading skill programs using the branching method of automated programed instruction. Two hundred inner-city fourth grade students were divided into two groups, low and average, on the basis of stanine scores of a previously administered achievement and IQ test. The experimental design was a 2 by 3 factorial with two levels of ability and three treatments. The three treatments consisted of two instructional procedures: subjects receiving training through automated reading programs; subjects receiving training in automated mathematics programs; and a control group in which the subjects did not receive automated instruction. The length of the experimental period was six weeks, with a testing period preceding and following that period. Each student in the treatment group worked in the Autotutor Center and was assigned materials three times a week for a twenty-minute period. The results tended to support the hypothesis that the use of automated reading instruction facilities learning for inner-city pupils. (Author/WR)