Four tufted capuchin monkeys (Cebus apella), successful in a tool task in which they used a stick to push a reward out of a tube, were tested in a similar task, with a tube with a hole and a small trap. Depending on where the stick was inserted, the reward was pushed either out of the tube or into the trap. With the trap-tube task, we assessed whether the monkeys understood the cause-effect relation between their behavior and the outcome. In Experiment 1, each subject underwent 14 10-trial blocks with the trap tube. Three subjects performed at chance level. The 4th subject's (Rb) performance improved, reaching 95% success in the last 6 blocks. In Experiment 2, Rb received additional tests to investigate its successful strategy further. Rb solved the trap-tube task by means of a distance-based, associative rule. The performances of the 4 subjects indicate that they did not take into account the effects of their actions on the reward.