Although the cesarean section rate has increased steadily for the past 12 years, further increase seems unlikely since the indications for performing the operation are already broadly defined. Most of the earlier indications will remain unchanged (eg, the presence of placenta previa and cephalopelvic disproportion). The trend toward vaginal delivery in perhaps 30% to 40% of breech births will probably have no material effect on the number of cesarean sections performed, and the present use of cesarean section for multiple pregnancy will probably continue. The two conditions under which cesarean section rates might become significantly lower are (1) automatic repeat cesarean section (which now accounts for more than 25% of all cesarean sections), a procedure which will probably decline as increasing numbers of such women have vaginal deliveries, and (2) a redefinition of the present midforceps classification, which will permit some of the easy midforceps deliveries from a low level to be performed without the legally abhorrent stigma of mid-forceps delivery. The value of prophylactic antibiotics for women predisposed to infection has now been proved, and further placebo studies to demonstrate this are not warranted. In the past, "type and match 2 units" was a routine prelude to cesarean section, and for every unit of blood transfused to cesarean section patients, some 25 units were cross-matched and held in (unnecessary) readiness. This formula is gradually giving way to type and screen, eliminating countless crossmatches. Because of possible harmful fetal effects, preoperative fluid loading, a necessary part of conduction anesthesia, is changing from the customary 5% glucose to the use of fluids containing no glucose. It has been suggested that conduction anesthesia may not offer unlimited time for cesarean section, as used to be thought. Apgar scores are lower if the time from uterine incision to delivery is longer than three minutes, a diminution that may be a function of the anesthesia or may reflect difficulty in delivery. Cesarean section mortality is much lower than it was in former years, but one may expect from one to two deaths per 1,000 operations. Overall, the maternal mortality from cesarean section per se is probably from three to five times higher than that of vaginal delivery (in one series, 11.5 times higher than vaginal delivery). The incidence of mild, transient respiratory signs in the newborn is higher after cesarean than after vaginal delivery, and the incidence of respiratory distress syndrome is also slightly higher.(ABSTRACT TRUNCATED AT 400 WORDS)