We studied primigravid women in spontaneous labor at term and given epidural analgesia. Two hundred such women giving birth in 1983 were compared with similar groups who gave birth in 1985 and 1987 after the introduction of ‘delayed pushing’ into our labor ward protocol. Four hundred controls were obtained in 1983 and 1985 by selecting from the labor ward register the next normal primigravid woman in spontaneous labor without epidural analgesia. the assisted delivery rate was significantly higher in all three epidural groups than in the controls (p < 0.001). Among epidural labors, there was no significant difference between 1983 and 1987 in the incidence of rotational or non-rotational forceps, or of cesarean section. In each epidural group, women who waited more than 60 min were less likely to experience a normal spontaneous delivery than were those who did not (p < 0.001 in 1983; p = 0.006 in 1985 p = 0.035 in 1987). We conclude that to delay pushing beyond 60 min confers no benefit regarding mode of delivery.