In animal research, blocking of instrumental responding is a well-known phenomenon, whereas no study has been reported that investigated blocking in human instrumental conditioning. Following the standard method used in nonhuman subjects, 48 students randomly assigned to one of three groups were exposed to a variable interval schedule (VI 10 s) in which reinforcement was delivered with a brief delay (500 ms). In the blocking condition, subjects experienced a tone stimulus during the delay (correlated group). In the two control conditions, subjects received either no tone (no-tone group) or the same number of tones as subjects of the correlated group, but the tones were independent of their behavior and reward (random group). As expected, instrumental responding was significantly lower in the correlated group than in either the no-tone or the random group. In a subsequent extinction phase, no difference in resistance was observed. A postexperimental interview revealed that subjects of the correlated group were more likely to detect the temporal nature of the reinforcement schedule than subjects of the other groups, but there was no relation to response rate. The data provide only little support for a notion of signal-induced enhanced learning, but do not challenge an interpretation in terms of associative competition between the response and the signal.