Two experiments tested both long-term retention of complex material by professional actors, and the contributions made by motoric codes to that retention. The first experiment indicated that, during performance, the actors rendered the written script with approximately 98% accuracy, making only very minor changes, primarily additions or deletions of verbal interjections. The second experiment assessed recall three months after the final performance, with the actors having learned and performed new roles in the interim. Retention was still extremely high with approximately 90% of the text being recalled verbatim or within very narrowly defined limits. A repeated trials paradigm (involving actors either sitting or moving about as they did in performance) indicated that motoric cues aided long-term verbal recall even when the movements did not constitute enactments of the literal words.