Since 1990, Uganda has hosted an estimated 200?000 refugees in postemergency settlements interspersed within host communities. We investigated the extent to which obstetric needs were met in the refugee and host populations during 1999-2002. Between September and December, 2000, we retrospectively collected data from 1999 and 2000 on major obstetric interventions for absolute maternal indications from all five hospitals in Arua, Adjumani, and Moyo districts, Uganda. The same data were collected prospectively for 2001. We did community-based maternal mortality surveys on refugee and host populations in Adjumani district in 2002. Rates of major obstetric interventions were significantly higher for refugees than for the host population who live in the same rural areas as refugees (1.01% [95% CI 0.77-1.25] vs 0.45% [0.38-0.52]; p<0.0001). Rates of major obstetric interventions were also significantly higher for refugees than for the host population who live in rural areas without refugees (1.01% [0.77-1.25] vs 0.40% [0.36-0.44]; p<0.0001). Maternal mortality was 2.5 times higher in the host population than in refugees in the Adjumani district (322 per 100000 births [247-396] vs 130 [81-179]. Refugees had better access to health services than did the rural host population in the northern Ugandan communities that we surveyed.