This paper describes two studies that were intended to replicate and extend the findings of Luke, Delanoy and Sherwood (2008), who were able to demonstrate a precognition effect using a covert task with contingent reward or punishment. Performance in their study was related to measures of belief in luck that could be considered to be related to experience of PMIR ‘in the field’. In Study 1, 25 participants completed the short-form Questionnaire of Beliefs about Luck (QBL: Luke, Delanoy & Sherwood, 2003) and a 10-trial preliminary preferences task that required them to select which of four fractal images they found most pleasant. In fact this was a precognition task and based on performance participants in the contingent condition subsequently either completed a pleasant task, involving rating cartoons for humorousness, or an unpleasant task, monitoring sequences of digits. Participants in the nocontingent condition completed neither. Overall, participants selected significantly more target images then mean chance expectation (t[24] = 2.60, p = .02), but there was no difference between the contingent and nocontingent conditions (t[23] = .73, p = .47). Performance was positively correlated with the Chance and Providence subscales of the QBL (r = .48, p = .02, and r =.39, p = .05 respectively) In Study 2, we added measures of openness to experience and creativity that we hypothesized to be related to PMIR performance as correlates of latent inhibition (LI) and lability respectively. 32 participants completed Goldberg’s (1999) measure of Openness to Experience, Holt’s (2002) Creative Cognition Inventory and Luke, et al.’s (2003) long-form QBL. All then completed the contingent version of the covert precognition task used in Study 1. Overall, participants again selected significantly more target images then mean chance expectation (t[31] = 2.01, p = .03). We did not replicate the correlations between performance and the Chance and Providence subscales of the QBL, nor with creativity measures, but there was a significant positive correlation with openness to experience, as predicted (r = .46, p = .01). Suggestions are given for further research utilizing this task, particularly in testing the assumption that the psi element need be covert