We conducted two case studies to evaluate the effectiveness of individualized remedial reading instruction linked to direct measures of student learning and transfer. The participants were two first grade children at serious risk of reading failure. The instruction targeted decoding of one syllable short vowel words and first emphasized onset-rimes and then phoneme manipulation. Forty words organized into 10 rime patterns were targeted for instruction. Probes consisting of all the instructional words were administered before each instructional session to monitor student progress. We also administered transfer probes to determine if the children could apply their knowledge more broadly. Instruction with onset-rime units yielded excellent maintenance and near transfer (novel words containing instructed rimes) but not far transfer (novel words derived from uninstructed rimes). The instructional change to phoneme manipulation yielded better far transfer for one of the two children. We consider the importance of administering measures that assess more than what is taught to get an accurate portrait of children's responsiveness to instruction.