Tests of the theory of morphic resonance have tended to confirm the theory’s predictions but are difficult to evaluate since they are typically reported in popular accounts rather than more detailed peer-reviewed journal papers. We replicated earlier work using word-based stimuli in a study that also looked at the effects of transliminality on performance. Sixty participants were exposed to five genuine Chinese characters and five false characters. Subsequently, participants identified the characters they could recognize among a sheet consisting of those originally presented intermixed with 10 decoys. As predicted, participants accurately recognized more of the genuine than false characters, t(59) = 2.40, p = .020, but also were more likely to report false memories for genuine than false characters, t(59) = 3.805, p = .001. Transliminality scores were related to performance with presented characters (r = .38; p = .003) but not with decoy characters (r = .14, p = .28)