Two experiments made use of a procedure known to generate latent inhibition in human associative learning. Participants received training consisting of exposure to a list of actions performed by a fictitious Mr. X. For most of his actions, an outcome was described, but some were not followed by any outcome. The last action performed by Mr. X was novel for participants in the NOVEL condition. For participants in the EXPOSED condition, Mr. X had performed that target action on repeated occasions, without it producing any outcome. After training, all participants were tested on their ability to retrieve what was the last action performed by Mr. X. In both experiments, retrieval of the target action was poorer in the EXPOSED than in the NOVEL condition. Experiment 2 also included a condition in which the target action was followed by a novel outcome and demonstrated a latent inhibition effect—poorer performance in the EXPOSED condition on a test of the association between the target event and its outcome. These results are interpreted in terms of an attention-reducing mechanism, triggered by the repeated preexposure to the target in the absence of a following event. It is argued that the attentional change involves a reduction in the effective salience of the stimulus of the target event, and thus reduces the processing necessary for encoding in memory and the ability of the event to enter into associations.