Background Women respond differently to anesthesia than men, initially recovering more rapidly, but having more postoperative morbidity. Studies on surgical patients report evidence of memory formation during anesthesia. However, sex differences in memory formation have not been explored. Therefore, the authors investigated sex differences in the implicit and explicit memory formation during general anesthesia. Methods With ethics committee approval, 120 consenting adult patients scheduled to undergo surgery during general anesthesia were recruited. Intraoperatively, 16 target words were presented to patients via headphones, and the Bispectral Index was recorded. Postoperatively, memory for presented words was tested using a word stem completion test. The test was divided into inclusion and exclusion parts, to delineate implicit and explicit memory contributions. Results Target and distracter hit rates were similar in men and women. For the whole study group, there was a significant difference between inclusion target hit rate (0.42) and base hit rate (0.39) (P = 0.01). Buchner's model suggested that this memory formation was attributable to both implicit and explicit memory. A Bispectral Index value greater than 50 was the only significant predictor of inclusion target hit rate. None of the patients were able to consciously recall the words presented during surgery. Conclusions Patients showed greater memory performance for words presented during general anesthesia than for words not presented. However, sex differences in memory formation were not observed. A relation between hypnotic state and memory during sevoflurane anesthesia was also established, suggesting that memory formation is possible even at hypnotic depths considered to be adequate anesthesia.