Fifty-two women undergoing labor induction and vaginal delivery at term were randomized between two oxytocin infusion protocols, involving hourly versus quarter-hourly increases in dose. Potential differences were sought for duration of labor, amount of uterine activity generated, and amount of oxytocin required. Starting at 0.5 mU/minute, oxytocin infusion was increased regularly in small increments every hour or every 15 minutes, according to group assignment. No differences were observed in potentially confounding clinical and demographic factors between the groups, including time to ruptured membranes. There were no clinically or statistically significant differences found for the duration of any phase or stage of labor, quantitative assessment of uterine activity, incidence of hyperstimulation, or neonatal outcome. The average dose of oxytocin used was lower in the hourly than in the quarter-hourly protocol (4.4 versus 6.7 mU/minute; P<.005). Significantly fewer patients on the hourly protocol required a maximum infusion rate exceeding 8 mU/minute (P<.05). More patients on the hourly protocol either had oxytocin discontinued completely or were maintained at 4 mU/minute or less during the active phase of labor (P<.05 and P<.001, respectively). We conclude that a slower rate of increase in oxytocin administration via continuous infusion results in no prolongation of any phase of induced labor, while permitting lower infusion rates of the drug.