Formerly Registrar, the City Hospital, Nottingham For many years general anaesthesia has been the anaesthetic of choice for most operative obstetric deliveries, but recent awareness of the danger of inhalation of vomit (Jeffcoate, 1953 ;Parker, 1954Parker, , 1956) has caused concern. The extent of the problem is revealed by reference to the Report on Confidential Inquiries into Maternal Deaths in England and Wales, 1952-4, which (Gordon, 1955;Martin, 1955 ;Barnes, 1955; Rollason, 1955 Although the safety of general anaesthesia has been increased by endeavours to reduce both the incidence of vomit and the risk of inhalation, complete success may not be achieved. The use of apomorphine (J. M. Holmes, 1956) There were no maternal deaths and, headache excepted, few maternal complications, none of which were serious. The results to the baby were also satisfactory. A "saddle block" technique has been used, 0.8 ml. of 0.5% hyperbaric cinchocaine hydrochloride (" nupercaine ") being given. This causes complete loss of sensation in the lower abdomen, perineum, and thighs. Sensation from the uterine cavity is greatly diminished; thus difficult forceps and breech deliveries requiring intrauterine manipulations can be performed without causing distress to the patient. There is complete perineal relaxation, which facilitates delivery and the performance of manipulations such as correction of the malrotated head in forceps deliveries and the bringing down of arms in breech cases.The conscious patient is able to assist delivery by use of her abdominal muscles; this is valuable in assisted breech deliveries. Nevertheless there is considerable loss of power and the perineal reflex is lost. It is therefore advisable to give the injection as late as possible in breech cases to avoid unnecessary extraction.Since uterine tone is retained under spinal analgesia the risk of post-partum haemorrhage is reduced, though intravenous ergometrine (0.25 mg.) has been given with delivery of the anterior shoulder in forceps cases and on completion of the second stage in breech deliveries. Post-partum haemorrhage (loss of 20 oz. (570 ml.) or more) occurred in 32 forceps deliveries and one breech (incidence 7.5%), but in 11 cases bleeding was thought to have arisen from cervical or vaginal tears. Manual removal of the placenta was performed in 13 patients (incidence 3 %); no undue difficulty was experienced.Forceps delivery was difficult in five patients who had mitral stenosis, and in three who had chronic bronchitis. Saddle block analgesia was considered especially suitable in these cases.