This study addressed three basic questions involving the amount of measurement needed to obtain valid estimates of reading growth. Participants were 41 students with learning disabilities from Grades 3 to 8, monitored across 10 weeks using curriculum-based measurement (CBM) of words read correctly (WRC). We compared growth slopes based on measurements taken weekly, every 2 weeks, every 3 weeks, every 4 weeks to an estimate of "true slope." Results showed that frequency of progress monitoring could be significantly reduced without detracting from the validity of growth estimates. However, validity was negatively affected by minimizing the number of WRC scores collected at each measurement occasion and by employing only one rather than four scores to estimate baseline. (Contains 3 tables.)