Investigated a hypothesis derived from attribution theory that a highly salient, task-contingent reward is more detrimental to intrinsic interest than a relatively nonsalient reward. In Exp I 60 preschoolers were either shown (salient-reward condition) or not shown (nonsalient-reward condition) a cue reminding them of the forthcoming reward. In subsequent immediate and delayed (4 wks) free-play periods, Ss in the salient-reward condition played less with the activity that had generated the reward than did Ss in the nonsalient or control (no-reward) conditions. In Exp II, 66 preschoolers were either asked to think about the reward, to think about an unrelated topic (distraction condition), or were not explicitly asked to ideate (nonideation condition). Subsequently, Ss in the distraction condition and a nonreward control condition displayed the most interest in the target activity. Results are discussed in the context of attribution theory and frustration theory.(27 ref)(PsycINFO Database Record (c) 2016 APA, all rights reserved)