Subjects designated as low and high psychotic-prone on the basis of MMPI subscales were divided into three instruction groups and then exposed to superimposed attended and unattended meaningless shapes. The instruction conditions were designed to influence the expectation as to whether the first phase (stimulus exposure) of the experiment would continue. One group was led to believe that there would be a second phase, another that there would not, and a third was given neutral instructions. In a subsequent recognition memory test, it was found that high psychotic-prone subjects had better recognition memory scores for previously unattended stimuli than the low psychotic-prone subjects in the neutral and continuation instruction conditions but not in the no continuation condition. These data are congruent with the findings that low but not high psychotic-prone subjects exhibit a latent inhibition effect. Both sets of data suggest that high psychotic-prone subjects process nominally unattended stimuli to a greater degree than low psychotic-prone subjects. Although the results are compatible with an explanation in terms of differences between low and high psychotic prone subjects in their strategies for allocating attention to the irrelevant stimuli, differences in the use of automatic and controlled information processes may be important.