Objectives The mode of delivery in diabetic patients is debatable. This study was designed to assess the pattern of delivery of macrosomic babies with a high prevalence of diabetes mellitus in Bahrain. Methods This retrospective analysis was conducted on mothers who delivered babies weighing ≥4.0 Kgs from 2001 to 2011 at Bahrain Defence Force Hospital. Data regarding patients' age, weight, mode of delivery, diabetic status, gestational age and parity were recorded. The main outcome was the effect of diabetes mellitus on the decision to allow vaginal delivery for macrocosmic babies. Other outcomes were failed trial of labour, parity, maternal age and foetal weight on the trial of labour and neonatal morbidity associated with vaginal births. Results The incidence of macrosomic babies was 2.2% of total births. Pre-existing diabetes mellitus was 3.9% of the study cohort. The rate of elective Caesarean section increased from 12.5% in non-diabetic mothers to 50% in patients with pre-existing diabetes. In cases of allowing a trial of labour, approximately 70% of patients with pre-existing diabetes had successful vaginal delivery. Patients with a previous delivery were less likely to undergo emergency procedures, but had the same probability for elective Caesarean compared with primigravida. Patient's age and foetal weight had no influence on successful trial of vaginal birth. Conclusions There was a trend to offer more elective Caesarean sections in patients with macrosomic babies in the presence of pre-existing diabetes. The majority of patients who were offered a trial of labour achieved vaginal delivery with minimal morbidity.