The present study tested whether participants would discount "won" money differently than they would "owed" money in a probability-discounting task. Participants discounted $1000 or $100,000 that they had won in a sweepstakes or that was owed to them using the multiple-choice (Experiment 1) or fill-in-the-blank (Experiment 2) method of collecting discounting data. When using area under the curve as the dependent measure, results from both experiments showed that participants discounted won money more than owed money and discounted $100,000 more than $1000. These results did not necessarily reflect what was observed at the extreme probabilities, however. The present results replicate those found with delay discounting of won and owed money. However, they question the idea that these differences are the outcome of owed money being more subjectively valuable than won money. Rates of discounting, be it delay or probability, appear to be at least partially determined by the outcome itself. Such a finding potentially complicates the issue of how probability discounting is related to delay discounting. (c) 2012 Elsevier Inc. All rights reserved.