The effect of task performance feedback and associated monetary earnings on drug self-administration were evaluated using eight subjects in a residential laboratory setting. The hypothesis was that if subjects believed that d-amphetamine impaired performance and reduced monetary earnings, d-amphetamine self-administration would decrease. Subjects performed computer tasks every day: on certain days that they received capsules, subjects were given bogus feedback regarding their performance ("better" or "worse" than average). On sample days, subjects were required to take d-amphetamine (10 mg BID) or placebo (0 mg BID) capsules. On choice days, subjects could choose between either d-amphetamine or placebo. Subjects received feedback on their task performance on 2 sample days and 2 of 4 choice days. Subjects received no feedback on the remaining two choice days. When subjects received no feedback, they chose d-amphetamine over placebo 78% of the time, and when they were given better feedback messages, they chose d-amphetamine 87.5% of the time. In contrast, d-amphetamine self-administration decreased significantly to 25% when subjects were told that it impaired their performance on work tasks and resulted in reduced earnings. In reality, d-amphetamine had little effect on work task performance. However, compared to placebo, d-amphetamine significantly increased subjective ratings of "Stimulated" and "Good Drug Effect" and significantly decreased ratings of "Tired" and "Sleepy." These results demonstrate that d-amphetamine served as a reinforcer under conditions in which drug self-administration did not influence monetary earnings, but that d-amphetamine self-administration could be modified by feedback/monetary earnings. Thus, contingencies associated with performance have important implications for drug use in the workplace.