Beyond reliability and validity, measures used to model student growth must consist of multiple probes that are equivalent in level of difficulty to establish consistent measurement conditions across time. Although existing evidence supports the reliability of curriculum-based measurement in reading (CBMR), few studies have empirically evaluated the appropriateness of the methods employed to determine the difficulty of CBM-R passages. The current study extends existing research with an evaluation of the relationship of oral reading performance and readability estimates on 50 CBM-R alternate forms. A sample of 88 second- and third-grade students read all 50 forms. Spache, Lexile, and Forcast readability estimates were used to estimate the difficulty of alternate forms. Results replicate and extend previous research to suggest that readability estimates are inadequate predictors of oral reading performance and therefore evaluation of CBM-R passage difficulty. (Contains 1 table.)