Objective: to assess the feasibility of obstetric offering community-led maternity care to most women, both those assessed to be at low obstetric risk and those with complicated pregnancies at 'booking'. Community-led care is defined as appropriate care by community midwives and general practitioners during pregnancy, birth and the puerperium, with routine hospital care kept to a minimum., Design: observational study., Setting: South Camden, London, UK: University College Obstetric Hospital and community., Participants: 453 women, resident in South Camden, including those 'booked' for home births, who were 'booked' for maternity care at University College Hospital between October 1993 and April 1994., Measurements and Findings: this paper assesses the extent to which community midwives and general practitioners were able to give local women community-led care and describes the amount of care provided to women by their 'named' community midwives and team. Most local women were eligible for community-led care and 85% planned to have it. The majority of care was given by the community midwives, but the amount of hospital input varied. Women who remained at low obstetric risk generally had their antenatal care in the community, only attending hospital for two or three routine assessments and occasional extra referrals. Women attending hospital more frequently usually had a complicated pregnancy. Care given by a woman's 'named' midwives was generally provided antenatally, but care from familiar midwives was less common in labour and postnatally., Key Conclusions and Implications for Practice: community-led maternity care can be provided to the majority of women, even those with a complicated pregnancy, as long as specialist opinion and facilities are accessible and women are referred as necessary. Although the majority of women had access to local antenatal care from staff they got to know, the 'named' community midwives and teams found it difficult to provide comprehensive care, particularly to the women who developed complications, so priorities need to be established. Further research is needed to compare alternative models of care and their costs.