We tested memory priming for auditory stimuli presented during general propofol-sufentanil anesthesia in 58 patients undergoing day-case arthroscopic surgery. Stimuli were presented via headphones and consisted of common facts (Group A, 29 patients), or familiar or unfamiliar full names of fictitious people (GRoup B, 29 patients). Group A was expected to give more correct answers to questions about the common facts than Group B, when tested postoperatively, and Group B to attribute more fame to presented names than Group A (famous names test). Because the process for learning new or unfamiliar stimuli (elaboration) in particular may be impaired under general anesthesia, more memory priming was expected for familiar than for unfamiliar material. No significant differences were demonstrated between the two groups in performance on common facts or in fame attributed to the names. The amount of memory priming, however, was positively related to one of two measures of preoperative anxiety.