A sample of 174 eleventh-grade students divided two passages into idea units (segments of information used when reading) and rated predetermined idea units according to three attributes that have been shown to correlate with recall. Standardized test scores for reading, scholastic aptitude, and general achievement were included in the analyses. Results of regression analyses and t-tests of differences between good and poor readers indicated that the distinctive skills of fluent reading do not involve the identification of idea units or the discrimination of idea-unit interest, importance, or concreteness. Correlations between standardized test scores and reading are also discussed. (Author/AA)