Selective attention was assessed in second, fourth, and sixth grade reflective and impulsive children with an incidental learning task using pictures (animal-household object pairs) or shapes (colored forms) as stimuli. By the sixth grade, reflective children displayed less incidental learning and greater central learning than impulsive children when picture stimuli were used. Also, reflective children showed a trade-off of incidental for central learning but impulsive children did not appear to attend selectively. There were no differences between these children in selective attention when shape stimuli were used. It was concluded that reflective children are more flexible in their deployment of selective attention. The findings were discussed in terms of Hagen's two-stage model of selective attention. The possibility that impulsive children are not adept at utilizing feedback to determine relevant and irrelevant components was discussed. (Author)