We extend to English Ito's (2001) demonstration of a hemispheric asymmetry in a false recognition and list-learning paradigm in Japanese. We show for English a significant interaction between the hemifield to which the probe word was presented and the probe type (non-studied but semantically related to the studied words versus non-studied and semantically unrelated to the studied words). The right hemisphere performed less well in rejecting the semantically related "lures"; the left hemisphere performed better in rejecting the semantically novel unstudied words. We discuss the results in terms of a model of fine semantic coding in the LH and coarse semantic coding in the RH, together with the dependence on memory probes that retain the representational signature of the hemisphere to which they were initially projected. Sex of the participants was also included as an independent variable; the results suggested that most of the effect lay with the female participants.