This was the first study that aimed to examine the effects of instructing participants to “fake” their performance on the Implicit Relational Assessment Procedure (IRAP). Thirty-six participants were first exposed to an IRAP. Consistent blocks involved responding to pleasant target words as pleasant, and unpleasant words as unpleasant; inconsistent blocks involved the opposite response pattern. As predicted, latencies were significantly shorter on consistent relative to inconsistent blocks. Subsequently, all participants were informed about how the IRAP works; 12 participants were also asked to try to fake the next IRAP by thinking of pleasant things as unpleasant and unpleasant things as pleasant; and another 12 were also told how to fake the IRAP (slow down on consistent trials). Results showed no evidence of faking, which contrasts with previous research in which the Implicit Association Test (IAT) was successfully faked.