Previous studies found that GBR 12909 can decrease cocaine-maintained responding at doses that do not affect food-maintained responding. In this study, the effects of GBR 12909 (0.3-3.0 mg/kg) were further examined by varying the response requirement and unit dose of cocaine. Rhesus monkeys earned food or cocaine under a multiple fixed-ratio (FR) schedule. The FR for food was always 30, but the FR for cocaine was varied from 10-130 and the unit dose was varied from 5.6-56.0 microg/kg per injection. Doses of GBR 12909 were tested in an ascending order, for 5 consecutive sessions each. GBR 12909 selectively decreased cocaine-maintained responding in all monkeys in at least 1 condition. These effects were enhanced with large response requirements and/or small unit doses. The results demonstrate that environmental variables can influence the selectivity of GBR 12909's effects and contribute to a growing debate concerning the evaluation of potential pharmacotherapies for drug abuse.