BackgroundA high percentage of abortions performed in South Africa are in the second trimester. However, little research focuses on women's experiences seeking second trimester abortion or the efficacy and safety of these services.The objectives are to document clinical and acceptability outcomes of second trimester medical and surgical abortion as performed at public hospitals in the Western Cape Province.MethodsWe performed a cross-sectional study of women undergoing abortion at 12.1-20.9 weeks at five hospitals in Western Cape Province, South Africa in 2008. Two hundred and twenty women underwent D&E with misoprostol cervical priming, and 84 underwent induction with misoprostol alone. Information was obtained about the procedure and immediate complications, and women were interviewed after recovery.ResultsMedian gestational age at abortion was earlier for D&E clients compared to induction (16.0 weeks vs. 18.1 weeks, p < 0.001). D&E clients reported shorter intervals between first clinic visit and abortion (median 17 vs. 30 days, p < 0.001). D&E was more effective than induction (99.5% vs. 50.0% of cases completed on-site without unplanned surgical procedure, p < 0.001). Although immediate complications were similar (43.8% D&E vs. 52.4% induction), all three major complications occurred with induction. Early fetal expulsion occurred in 43.3% of D&E cases. While D&E clients reported higher pain levels and emotional discomfort, most women were satisfied with their experience.ConclusionsAs currently performed in South Africa, second trimester abortions by D&E were more effective than induction procedures, required shorter hospital stay, had fewer major immediate complications and were associated with shorter delays accessing care. Both services can be improved by implementing evidence-based protocols.