This exploratory study identified underlying skill and ability differences among subgroups of adolescent and young adult struggling readers (N = 290) overall and in relation to a fluency-based instructional grouping method. We used principal axis factoring of participants' scores on 18 measures of reading-related skills and abilities identified in the research literature to identify a smaller set of generally uncorrelated constructs. The four underlying factors of the 18 measures explained 62.7% of the variance. We labeled these factors Encode/Decode (44.5%), Vocabulary (9.5%), Processing Speed (5.2%), and Working Memory (3.5%). Regression analysis demonstrated Working Memory, Encode/Decode, and Vocabulary collectively predicted 45.9% functional reading as measured by the "Comprehensive Adult Student Assessment System." Alternatively, when measured by the "Test of Adult Basic Education," Vocabulary and Encode/Decode predicted 47.1% of variance in reading. Differences in predictive utility of the factors by fluency group suggest approaches to tailoring instruction for each group. Future research might examine the optimal mix of instructional approaches that support the identified factors.