Two student self-management techniques, student charting and student selection of instructional activities, were applied to ongoing data-based program modification. Forty-two elementary school resource room students were assigned randomly (within teacher) to one of three treatment conditions: Teacher Chart-Teacher Select Instructional Activities (TC-TSA), Student Chart-Teacher Select Instructional Activities (SC-TSA), and Student Chart-Student Select Instructional Activities (SC-SSA). Statistical analyses indicated that the SC-SSA group improved on all three dependent measures of words read correctly from passages, whereas the TC-TSA and SC-TSA groups each improved on two of these. Also, the SC-SSA group was the only treatment condition that improved on the standardized reading test measure. Implications for making substantive changes in instructional plans are discussed. (Author)