Prolactin and somatotrophin were measured during the postoperative period in two series of 15 patients after gynaecological surgery. Samples were collected for four days at the same times during the 24 hours period. The anesthetic given in the first group was a neuroleptanalgesia of dextromoramide-droperidol type followed by postoperative analagesia using a noramidopyrine compound. In the second group, epidural anaesthesia was given, followed postoperatively by the injection of lidocain at constant rate interrupted between the final two samples. In the neuroleptanalgesia group, from a basal levels of 11 micrograms.l-1, prolactin rose to 22 micrograms.l-1 on the evening after surgery (p less than 0.001) to subsequently stay on a plateau between 6 and 8 micrograms.l-1 (p less than 0.025 to p less than 0.005). From a basal level of 2.8 micrograms.l-1, somatotrophin rose to 9 micrograms.l-1 (p less than 0.05) then fell progressively from 7.5 to 2 micrograms.l-1 (NS on D1, D2, D3). In the epidural group, from a basal level of 13.5 micrograms.l-1, prolactin rose to 23 micrograms.l-1 on the evening after surgery (NS) to fall sharply on D1 to 5.6 micrograms.l-1 (p less than 0.01) and then follow a plateau on D2 and D3 of the order of 11 to 12 micrograms.l-1 (NS). From a basal level of 1.9 micrograms.l-1, somatotrophin rose to 10 micrograms.l-1 (p less than 0.001) to fall again to 4.5 micrograms.l-1 on D1 (p less than 0.01) and to 2 micrograms.l-1 on D2 and D3 (NS). Comparison of these two groups showed a difference only on D2 with regard to somatotrophin (p less than 0.05) and on D2 and D3 with regard to prolactin (p less than 0.025 and p less than 0.05). These results are discussed. They do not indicate any fundamental difference in the endocrine response to aggression in relation to the two types of anaesthetic studies.