This study evaluated delay discounting functions of substance abusing problem gamblers, substance abusing non-problem gamblers, and non-problem gambling/non-substance abusing controls. Subjects chose between hypothetical monetary amounts available after various delays or immediately. In one condition, a USUS$1000 reward was delayed at intervals ranging from 6 h to 25 years. At each delay interval, the immediately available rewards varied from USUS$1 to USUS$999 until choices reflected indifference between the smaller immediate and larger delayed rewards. In a second condition, the delayed reward was USUS$100, and immediate rewards varied from USUS$0.10 to USUS$99.90. In all three groups, hyperbolic discounting functions provided a good fit of the data, and the smaller reward was discounted at a higher rate than the larger reward. Substance abusers discounted delayed rewards at significantly higher rates than non-substance abusing controls, and problem gambling substance abusers discounted delayed rewards at higher rates than their non-problem gambling substance abusing counterparts. Discounting rates were not associated with types or recency of drug use. These results provide further evidence of more rapid discounting of delayed rewards in substance abusers, and especially among substance abusers with a comorbid gambling problem. Rapid discounting of delayed rewards may be a feature central to impulse control and addictive disorders.