This study contrasts the validity of 2 early reading curriculum-based measurement (CBM) measures: word identification fluency and nonsense word fluency. At-risk children (n = 151) were assessed (a) on criterion reading measures in the fall and spring of first grade and (b) on the 2 CBM measures each week for 7 weeks and twice weekly for an additional 13 weeks. Concurrent and predictive validity for CBM performance level and predictive validity for CBM slopes demonstrated the superiority of word identification fluency over nonsense word fluency. Findings are discussed in terms of the measures' utility for identifying children in need of intensive instruction and for monitoring children's progress through first grade.