We examined the effect of a teaching method on skill fluency and on-task endurance of a 9-year-old boy who had been diagnosed with attention deficit hyperactivity disorder. An academic task that occurred at low fluency during 10-min baseline sessions was taught to fluency. When responding was not yet fluent, brief reversals to baseline showed that the learner's rate of responding decreased and that he did not spend entire sessions on task. However, once a fluency goal had been reached, responding remained fluent and he remained on task in the third reversal condition.