BRITISH MEDICAL JOURNAL 31 JULY 1971 279 that prostaglandin E2 is more effective in inducing labour must be considered with caution, for though the patients were selected at random the groups were not identically matched. The differences found in these studies may be partly explained by essential differences in the plan of each investigation and the exact protocol undertaken. Beazley and Gillespie (1971) started the infusion of prostaglandin E, at 0-21 [Lg/min and oxytocin at 2 1 mU/min. These rates were doubled every hour until contractions were effective and then usually maintained at a fixed rate. In addition some patients' membranes were ruptured artificially when labour was established. Our approach was basically different in that rapid titration regimens were used to induce an early onset of regular contractions in each group, as we considered it was impossible to know that the stimulatory effect of doubling the infusion rates of prostaglandin E2 was quantitatively equivalent to doing the same with oxytocin. It is therefore not surprising that the results were different.By using the titration regimens an initial uterine response occurred at a comparable time with both substances, though regular contractions of similar amplitude generally occurred earlier with oxytocin. Despite this fact prostaglandin E2 was more effective in dilating the cervix, and it is our impression that this substance had a greater stimulatory effect on the uterus than oxytocin in the dose range used. One feature tending to confirm this impression was the finding that while prostaglandin E2 invariably resulted in spontaneous rupture of the membranes, this occurred before dilatation of the cervix had reached 6 cm in eight out of 15 patients. With oxytocin this was so in only two out of the nine successes. With both agents basically two types of response were elicited. One resembled normal labour in that progressive cervical dilatation occurred with time, while in the other little or no change in dilatation resulted for many hours, usually until the onset of spontaneous membrane rupture. About one-third of the successes in each group were in this latter category.With regard to the practicality of using oxytocin to induce labour in the presence of intact membranes, we feel that a greater success rate could probably be achieved if the infusion rate were increased above that required to induce contractions which appear "optimum" on an external tocograph. In effect oxytocin titration would be against cervical dilatation rather than contractions. The limitation of such a policy is a possibility of causing uterine hypertonus, though oxytocin infusions in doses ranging from 2 to 128 mU/min. We feel, however, that there are possible dangers of pursuing a similar policy of increasing the dose of prostaglandins above that required to induce optimum contractions. Besides maternal disturbance caused by possible vomiting and soreness of the arm it is our impression that fetal distress may be provoked, for contractions of excessive frequency or str...