In a retrospective study, 155 women who had epidural analgesia only for pain relief during labor and delivery were matched by age, parity, and race with 155 women who had either pudendal or local analgesia for delivery. Infant birth weights were similar in the two groups. The frequency of low forceps, vacuum extraction, and midforceps delivery in the epidural group was significantly higher than in the control group (P less than .05). Women having epidural analgesia who delivered their infants in an occiput anterior position also had a significantly higher frequency of instrumental delivery. Thus, epidural analgesia increased the frequency of instrumental delivery, an increase that was not a consequence of larger infants or fetal malposition.