The ephemeral reward task involves providing subjects with a choice between two distinctive stimuli, A and B, each containing an identical reward. If A is chosen, the reward associated with A is obtained and the trial is over. If B is chosen, the reward associated with B is obtained but A remains, and the reward associated with A can be obtained as well. Thus, the reward-maximizing solution is to choose B first. Although cleaner fish (wrasse) and parrots easily acquire the optimal response by choosing B, paradoxically, several nonhuman primate species, as well as rats and pigeons, do not. It appears that some species do not associate their choice and reward with the second reward. Surprisingly, research in an operant context with pigeons and rats suggests that inserting a delay between the choice and reward facilitates optimal choice. It is suggested that impulsivity may be, in part, responsible for the difficulty of the task. In an attempt to better understand this task, we trained human subjects on an operant version of this task, with and without a brief delay between choice and reward and found that many subjects failed to learn to choose optimally, independent of the delay. Furthermore, performance on this task was not correlated with a task thought to measure impulsivity, the Balloon Analog Risk Task or with the Abbreviated Impulsivity Survey. We concluded that, for humans, the task is confusing because there is no incorrect response, only good and better, and better is not easily discriminated. (PsycInfo Database Record (c) 2024 APA, all rights reserved).