People generally prefer an immediate reward to a delayed reward of the same amount. Such preferences can be understood in terms of temporal discounting. Most discounting experiments measure delayed reward value by having participants make a series of choices between an immediately available reward and a delayed reward. The choices are frequently hypothetical. Use of hypothetical rewards in most delay discounting experiments calls into question the validity of the results. Real rewards, which are of true interest, may produce results different from those of hypothetical rewards. The present experiment used a within-subject design to measure delay discounting of real and hypothetical rewards, and employed a range of real reward magnitudes that exceeded those used in previous delay discounting experiments, permitting a systematic test of a well-established finding. Results suggest that real rewards are discounted to a greater extent than hypothetical ones.