OBJECTIVE: This study aimed to investigate pregnancy outcomes in Somali-born women compared with those women born in each of the six receiving countries: Australia, Belgium, Canada, Finland, Norway and Sweden. DESIGN: Meta-analyses of routinely collected data on confinements and births. SETTING: National or regional perinatal datasets spanning 3-6 years between 1997 and 2004 from six countries. SAMPLE: A total of 10 431 Somali-born women and 2 168 891 receiving country-born women. METHODS: Meta-analyses to compare outcomes for Somali-born and receiving country-born women across the six countries. MAIN OUTCOME MEASURES: Events of labour (induction, epidural use and proportion of women using no analgesia), mode of birth (spontaneous vaginal birth, operative vaginal birth and caesarean section) and infant outcomes (preterm birth, birthweight, Apgar at 5 minutes, stillbirths and neonatal deaths). RESULTS: Compared with receiving country-born women, Somali-born women were less likely to give birth preterm (pooled OR 0.72, 95% CI 0.64-0.81) or to have infants of low birthweight (pooled OR 0.89, 95% CI 0.82-0.98), but there was an excess of caesarean sections, particularly in first births (pooled OR 1.41, 95% CI 1.25-1.59) and an excess of stillbirths (pooled OR 1.86, 95% CI 1.38-2.51). CONCLUSIONS: This analysis has identified a number of disparities in outcomes between Somali-born women and their receiving country counterparts. The disparities are not readily explained and they raise concerns about the provision of maternity care for Somali women postmigration. Review of maternity care practices followed by implementation and careful evaluation of strategies to improve both care and outcomes for Somali women is needed.