This study investigated the functional relationship between student accurate response levels and two mathematics drill procedures using a BCBC across participant's single case research design (in this case, "B" represents peer drill and "C" represents computer drill). Each of four elementary school students was randomly assigned to one of two dyads and presented addition flashcards under a computerized drill condition and a peer-mediated drill condition. In both conditions, students were drilled for 3 minutes on addition facts with immediate feedback (i.e., "Correct" or "Not correct"). Results suggested that the two somewhat older students showed higher levels of accurate responding in the computer condition, while the two younger students showed higher levels of accurate responding in the peer-tutoring condition. Discussion focuses on potential hypotheses for performance discrepancies across dyads, importance of matching instructional conditions with idiosyncratic variables, and implications for practice and future research.