Delay discounting was investigated in opioid-dependent and non-drug-using control participants. The latter participants were matched to the former on age, gender, education, and IQ. Participants in both groups chose between hypothetical monetary rewards available either immediately or after a delay. Delayed rewards were $1,000, and the immediate-reward amount was adjusted until choices reflected indifference. This procedure was repeated at each of 7 delays (1 week to 25 years). Opioid-dependent participants were given a second series of choices between immediate and delayed heroin, using the same procedures (i.e., the amount of delayed heroin was that which could be purchased with $1,000). Opioid-dependent participants discounted delayed monetary rewards significantly more than did non-drug-using participants. Furthermore opioid-dependent participants discounted delayed heroin significantly more than delayed money.