The purpose of this study was to ascertain the effects of manipulating the data base used for instructional decision making on student achievement. The dependent data were the number of words read correctly by 117 students on three basal reading passages. Students were rank ordered by the degree of implementation of a technically adequate curriculum-based measurement and evaluation system. The top 27% were compared to the bottom 27% by means of t tests on the scores. A significant difference was found between high and low implementation groups on one reading passage; differences between scores on two other passages were not significant but were in the same direction, supporting the hypothesis that a high degree of implementation of a technically adequate data system does lead to greater achievement. The discussion addresses implications for teaching practices.