ABSTRACT: Objective: The study aims to determine the differences in maternal and perinatal outcomes between caesarean and vaginal deliveries and the factors affecting vaginal delivery in twin pregnancy. Materials and methods: An observational study to audit twin pregnancies delivered at the University of Nigeria Teaching Hospital between 2002 and 2008. Clinical observations were entered into a questionnaire immediately after the delivery of the women and the mothers and their babies were followed up until the end of the puerperium. Results: There were 5298 deliveries within the study period, out of which 117 were twin deliveries. This gives a twinning rate of 22 per 1000 deliveries. The mean age of the mothers was 30 ± 5.9 years. Twenty-five (21.4%) women were admitted into the hospital for preterm labor. The average gestational age of admission was 32 ± 5.8 weeks and the average duration of hospital stay for preterm labor was 12.6 ± 9.1 days. Other pregnancy complications observed were severe hypertension (14.5%, anemia (9.4%), postpartum hemorrhage (8.5%), puerperal fever (5.1%), abruptio placentae, and diabetic mellitus (2.7%). Fifty-nine women (50.4%) had vaginal deliveries, 5 (4.3%) had vaginal delivery of the leading twin and caesarean delivery of the retained second twin while 53 women (45.3%) were delivered by caesarean section. Forty-eight (41%) women had preterm delivery. Vaginal deliveries were more common than caesarean section among patients that were unbooked than booked P = 0.047 (OR 2.26, 95%CI:0.93-5.53) and those that had cephalic presentation of the leading twin, P = 0.0002 (OR = 4.7 95% CI:2.6-8.2). Vaginal delivery tended toward statistical significance when the fetal weight of the leading twin was 1.5 to 2.5 kg, P = 0.09. The commonest indications for caesarean section were abnormal lies and presentations and hypertension in pregnancy.Two-hundred and seventeen (92.7%) out of a total of 234 fetuses that were delivered in this study were live births and 17 (7.3%) still births. The rate of new born admissions in twin 1 was however higher in those delivered by Caesarean section (39.6%) than those delivered vaginally (29.7%). Indications for admissions into the special baby care units were; prematurity 33 (40.2%), birth asphyxia 15 (18.3%), low birth weight 12 (14.6%), neonatal jaundice 10 (12.2%), and twin-twin transfusion 4 (4.9%). There was a higher rate of early neonatal death in both vaginally delivered twin 1 (9.4%) and twin 2 (11.9%) than those delivered by Caesarean section, 3.8 and 3.5%, respectively. Conclusion: Cephalic presentation of the leading twin, birth weight less than 2.5 kg, and unbooked women presenting in advanced labor predisposed to vaginal delivery in twin pregnancies. There was however increased risk of still birth and early neonatal deaths especially for the leading twin in vaginal deliveries in unbooked women.