Adverse effects of material rewards on Wechsler subscale performance may be the result of a reward-produced developmental regression. To further explore that idea through replicating earlier findings with adults, and to extend the enquiry to children, selected Wechsler subscales were administered to 32 subjects at each of three ages (5, 10, and 18 years) under either reward or nonreward conditions. Subscales were chosen to represent both algorithmic and heuristic types of tasks. Reward and nonreward groups (eight males and eight females per group at each age) were matched initially on age and ability. For adults, consistent with earlier findings, reward had an adverse effect on performance on the heuristic subscales and tended to facilitate performance on the algorithmic subscales. However, rewards generally had no effect at the fourth grade level and a reverse effect at the nursery school level, i.e., rewards facilitated heuristic and hampered algorithmic performance. These same trends were evident at both higher- and lower-ability levels, assessed via median split, except that at the college level, reward facilitation of algorithmic performance occurred only at the lower ability level, and detrimental effects of reward on heuristic performance were more pronounced at the higher ability level. Also, at the fourth grade, reward had an adverse effect on the algorithmic subscale performance of higher ability subjects. These complex findings appeared to be more consistent with an explanation based on developmental regression than on any available alternative mechanism. (Author)